NEAR Protocol Town Hall - July 2025

Captions as YouTube publishes them, usually machine-generated, so product and people names can be misheard.

1. 0:16

   GMGM everyone. I'm Peter from New Week, your source for what's next on NIA and welcome to the NIA town hall July edition. The world is in a moment of deep transition. The internet is changing. First it was pages, then apps, now it's agents. AI agents won't just write essays or generate images. They'll act, trade, make decisions on our behalf, on their own. And to make that work, we need more than just smarter models. We need infrastructure. We need

2. 0:39

   railroads. We need something to make open source AI monetizable, owned by its users, and help millions, eventually trillions of agents coordinate. That is near protocol. So when we say NAS for AI, we mean it, and we're building it.

3. 0:54

   Let's get in to the town hall. First up, as always, Nia's co-ound co-founder and the guy who's helped put the T in GPT, Ilia Pulukin. Welcome, Ilia. Stage is yours. Hi, thank you, Peter. Hey, everyone. I hope you're having amazing new summer.

4. 1:16

   Uh, it's been quite an eventful few past weeks. Let's go to the next slide. Um so there's obviously been a number of events that have happened uh from New York to K to Paris uh where we've been present. We had open source AI summit at permissionless as well as lots of activations with different groups institutionals founders um developers.

5. 1:47

   There was uh hackathons and then the kind of presence and again everything from debate of rollups versus sharding uh to introducing the road map to us AI and then in Paris engaging with uh kind of actual AI community um and uh explaining to them why crypto is needed.

6. 2:11

   I think the marketing team will dive in more into some of the events and highlights. Um but I think the high level is uh we obviously have a lot of people in the industry that realize that confidential computing confidential AI is extremely important and long-term making sure that ownership of that AI and ability for it to transact are key elements the future.

7. 2:43

   uh we'll go through some of the team updates and progress ecosystem wins and I want to kind of continue reemphasizing that we have the fork set campaign going and I want everyone to continue sharing it. So let's go to the next slide.

8. 2:57

   So there's few changes we've had at Near Foundation. So one I'm really excited to uh welcome Bowen uh to join Near Foundation console. Bowen has been with Near since day one when we kind of pivoted from AI to blockchain. He actually joined and interviewed for near AI originally. Um but joined and uh learned and kind of effectively built and then been leading the protocol team uh since and been effectively amazing leader for the space and for Neo ecosystem overall.

9. 3:35

   And so I want to welcome him to the NEA Foundation Council. uh effectively becoming not just kind of technical leader but also uh oversee the growth and direction of where Neo foundation is investing and putting its resources. We also have George Zang uh joining Neo Foundation as chief growth officer.

10. 3:58

    George previously was kind of founder as well as an operator across crypto and web two projects. He's been a product leader at Meta, has been a COO at UIDX and so bringing an amazing uh kind of uh scope of experience across kind of product growth leadership and it's been extremely welcomed by the team as as he's ramping up. So I'm sure you'll hear a lot more from him in coming months. Uh we also want to say thank you to Chris

11. 4:36

    Dorovan who's been transitioning from his full-time NF role to uh member of NFC. He's been on NFC for the past uh year and going to continue serving in that capacity and uh want to say amazing um kind of work and uh thank you for the for the contributions and then we have new amazing leaders joining this summer.

12. 5:03

    They've been really focused on bringing exceptional people to the ecosystem and to their foundation and so stay tuned on as they uh join the board. All right, next slide. Yeah, so uh we've kind of foundation operates under the OKR model objective and key result. um we have uh kind of per quarter and this is a summary for the first half right we're kind of through the half of the year um

13. 5:36

    and there's effectively a flywheel that's operating um that kind of how we're modeling the whole ecosystem where uh you know you bring users and you know create demand that uh leverages all the ecosystem applications wallets um as well as kind of overall infrastructure and social layer and then you know build the momentum bring the products new products existing products grow them and then kind of use those

14. 6:10

    products to bring more users and make them aware about and excited about so this is kind of a flywheel obviously there's founders coming in and uh contributing to this there's uh people you know token holders it's kind of all this role goals are kind of embedded into this firewall. Now I wouldn't say that um we've you know out executed every single goal but we are trying to put a very ambitious goals as well. So um for example you can see on the product side we actually consider that we missed our ambitious goals even

15. 6:43

    though intents done exceptionally well and been doubling every month since the beginning of the year uh including uh in June but we have put a very kind of for Q2 we had a very ambitious goal and we have missed it so that's kind of the model we're using so this is just for public accountability as we are kind of continuously striving to do better across the board but also put very ambitious goals to drive us forward.

16. 7:16

    Next slide. So, uh I'm sure you saw but we just had a Bitwise near ETP launched on Frankfurt Stock Exchange. It's been a pretty amazing um kind of event for the Neo Foundation team, Bidwise really coming together and bringing what effectively is a kind of blockchain native asset and uh product right to all the traditional markets that are operating through through these exchanges. which is exciting because you know as we're rolling out Near kind of

17. 7:59

    as this future AI currency as this instrument that AI and people will use to own um kind of in the future um this is actually opens up more doors for this and creates kind of bridges into more traditional finance. So really really exciting. I want to thank the team. I want to thank the wise. There's a lot of work to get this going and I'm sure this is not the last instrument like this and we're going to continue to see more and more of this integration happening. Next slide.

18. 8:39

    Um I mentioned Neo intents. We had a very ambitious goal but even that um they've been growing extremely well. Um over 102 assets supported. A number of new chains were added. There's even more chains in a pipeline right now. Uh there's actually some that are already in interface that you may find.

19. 9:01

    The number of swaps is uh kind of approaching a million which again really exciting but also we need a lot more. And the volume uh as I said been doubling every month. So like this past 30 days volume effectively was as much volume as we had before that. So to 10 million and so exciting to see also alltime high was 25 million per day. So I think we have you know we have a number of interfaces that integrated

20. 9:32

    source swap kyber swap obviously infinex um a number of others is also digital ecosystem uh I think the search uh swap just launched engineer and integrated obviously hot and others so it's kind of exciting to see this protocol really becoming this robust uh liquidity layer that we envision it be and um you There's a lot of work in progress to continue generalizing it and adding

21. 10:01

    ability to have natural language intents and also just generic escro style where solvers can take time to actually fulfill it. Uh creating even better liquidity utilization. Next slide. Uh I also just want to highlight uh hot team has launched an NFT marketplace and trading facility that is chain abstracted and kind of intentbased. And so this is really great example of ability to leverage now the intent as kind of this fundamental building block to connect with wallets across all

22. 10:40

    chains to connect with liquidity from all chains and then really build applications that don't have boundaries. And so I'm I'm excited to see I mean there's kind of some upcoming um announcement there but uh even the current authority came out right is about public AI maintaining their databases and so I think kind of the fundamental to me is you know if you want to build applications now in web 3 not just on

23. 11:10

    mirror you know be that you want to target salana users or you want to target base users or telegram users or um you know hyperlquid users. Actually, Near is the place to do that because you can actually open up your uh kind of addressable market by allowing all those users to come in, use their assets, use their wallets and interact with application and not even think about the bridging and you know changing wallets and all this kind of logistics that historically been needed. Next slide.

24. 11:47

    Finally, there's a lot of things happening on uh AI side. I will not go into kind of all of them, but just to highlight, proximity released the shade agent sandbox. It's effectively a very simple way how to go from a JavaScript implementation to a fully rolled out shade agents, right? where your brain of agent runs inside just execution environment in verifiable way and then your muscle right is a smart contract running on chain executing actions being able to transact being able to do things

25. 12:18

    on chain public AI secured 10 million in funding we build a human layer that data labeling crowd sourcing leveraging uh human intelligence to train AI and improve the models and and so Obviously they're partnering with hot as I mentioned but also exciting to see them grow and continue expanding and offering this to more and more companies especially as we see the AI market right kind of had a consolidation of scale.

26. 12:51

    Silverstream which is browser use company that works very closely with enterprises uh have achieved the breakthrough kind of accuracy on enterprise automation benchmarks. Uh so this is really about you know automating boring boring parts of jobs and enterprises and really allowing AI to take over the computer and do some of the things while all of this is accessible to everyone in easy way.

27. 13:19

    Finally beta uh launch their integration is intense which is really cool. Now again anybody can trade anything uh just by typing hey I want you know dodge to Avac or BTC to soul and it uses intense to actually integrate kind of solve that but also the bit itself got integrated into rhea and vex dexes creating AI system that really can guide you not just through trading but also help with a lot of different operations on those applications I think that's a really

28. 13:53

    important kind of milestone because that kind of shows the future of um changing experience how we're going to kind of merge the user interface and kind of this AI experiences together into a single unified kind of super app experience.

29. 14:13

    Next slide. So just kind of to conclude some of the learnings. I think near intense has made a very clear proof point for near I think there's a lot of just like ability to partner with different ecosystems and bring their folks together on on this journey while you know removing some of this like animosity that that three ecosystems and tokens usually have. I think to me there's been a very kind of big um both lesson and also relief in a way because I always felt kind of

30. 14:46

    strange when we needed to fight with different ecosystems instead of really kind of bring them all together. Um I think we clearly see that market wants more AI proof points. The AI space is moving so fast. There's so many things happening and we need to continue showing what what user own AI can do and why that is important. Um there's been a constant feedback around developer documentation making sure developers are staying able to find there's a lot of things

31. 15:18

    happening on near obviously chain signatures intents AI stuff the blockchain itself you know JavaScript python there's a going go uh to go smart contract so making sure uh so dev hub just released new documentation but you know that's definitely one of the big areas of feedback and then continue to recruit top talent I across the board across the ecosystem.

32. 15:41

    We need to continue bringing you know the best founders and the best talent across teams. Uh near intents definitely need more liquidity and there's been few stability issues uh over the past few months. Uh so improving that is a big area supporting D5 projects. I think we have a lot of really cool really promising projects but it's been pretty slow to come to market. I'm really excited to see Templar. I'm really excited to see fellow land and others um

33. 16:12

    and so we need to continue supporting them kind of with this journey to main net right I mean templar is in beta already finally we need more adoption on the AI project side right and clear GTM I think there's a lot of really cool experiments there's a lot of cool ideas and uh we need to really bring them to market so big area for improvement so the QC priorities are pretty simple. Grow near intense, you know, volume, uh, more

34. 16:42

    users, more integrations, uh, drive capital formation for projects in the ecosystem. Uh, really, you know, everything from fundraisers from VCs to token launches, making sure there's like continuous way to capital form, you know, leverage intense now where the token launches can be omniresent and anyone from anywhere can participate.

35. 17:05

    And finally, adoption of near exist projects really bringing them to market and working with them on clear DTM. Next slide. Cool. Uh for those who haven't seen, we have an inflation reduction proposal from the ecosystem. Uh hot and linear protocol have collaborated on this. If you haven't participated or voted, u this is a really great time and opportunity. um if you are against uh please you know communicate as well uh

36. 17:39

    and uh I think it's it's a really great that we have the community kind of coming together we have you know obviously in collaboration with protocol team uh working on all of this and kind of coordinating this this approaches there's also I just saw today there's a house of stake test kind of the first um MVP of the front end So also excited to see some of the governance work there uh which I think will you know continue using uh that methodology going forward. Next slide.

37. 18:16

    Finally uh just keep uh reinforcing the message that we're this fork of the road. I mean we just had Grog 4 coming out. It's beating all of the uh benchmarks. it's clearly advancing toward you know being more intelligent at least in smaller tasks and like specific types of tasks than uh kind of single person can be and so I think it's it's becoming more and more important and relevant that we actually make sure

38. 18:47

    that this types of models are neutral community and userowned versus you know potentially biased in all kinds of ways. So, uh, really important to spread this message to, you know, get more people and more talent to join the movement because that's how we're actually going to make this successful. Thank you.

39. 19:14

    Thank you so much for coming on and everyone. I encourage you as well to go vote on the inflation proposal as well as spreading the word about the fork that campaign. Share your fork that moments. Next up, we got our very own Jordan Gray from Public AI, a near OG, and he's here to tell us a little bit about what he's been up to lately with Public AI. Jordan, it's all yours.

40. 19:46

    Thanks for having us on. Um, yeah, great to be here. Appreciate the shout out for Millia earlier regarding our fundraising round that we closed out. So going to take this opportunity to take everybody through public AI as a whole in case um you're not familiar with what we're up to and then also show some of the uh current activities that people can participate in since we're live with those. So public AI at large um came out of this problem that AI is hungry for

41. 20:18

    data. Um data being valuable isn't something new. It's been um valuable in terms of marketing and uh everything else. Uh data brokerages are a healthy industry um to this day. But AI in specifically brings up another huge use case for data. And um if you look at what's really holding AI back, um things like compute, memory, physical resources, those are of course bottlenecks that hyperscalers are the only ones that can take advantage of to be able to do foundational model training. But as more and more

42. 20:55

    innovation happens, those um compute resources and so on become more and more commoditized. And the thing that is still really hard to get is valuable data and especially niche on demand data. Um after you've gotten a foundational model or something built, fine-tuning it to specific business use cases requires um a lot of like new data that's not available on the internet. Everybody's already

43. 21:25

    scraped the internet. Um synthetic data only gets you so far and then leads to model collapse in the final training. And to really take things to the next level, we need fresh on demand data. That's why things like X are valuable and Reddit etc. So what we've done is take a web3 approach to how to gather that data rather than a centralized approach. And what that looks like um is a maybe complex looking chart. So we can go over this uh layer by layer.

44. 21:57

    At the very top we have our clients that come to us looking for this niche data to do model training and so on. And then below that we have our kind of like product level or platform level where we have the data hub which is a DAP where people can come and do the data sourcing the uploading verification of that data and then also participate in rating models after the data has been applied.

45. 22:22

    So um human in the loop feedback for how the training is going and then we have data hunter which is our chrome extension which can be used for um other kinds of data fetching and then also offloading the compute that's required for the mechanical aspect of the data verification that happens and then below that is where blockchain comes in. So when you're talking about gathering high-quality data from a diverse group of contributors, you also

46. 22:51

    get diverse quality. So we need ways of identifying not only that high-quality data but high-quality contributors. And what that looks like is an onchain reputation system that uh tracks how people have behaved in the system, their voting patterns, etc. making sure that they are making meaningful contributions. Um we've got a consensus voting mechanism in place and that's the data verification

47. 23:14

    uh element and it operates a lot like a DAO um as far as how that data gets validated or even um similar to how a blockchain uh validates its blocks. After a certain number of people all say yes, this is valid good data, then it goes um off to the data set for delivery to client. And to ensure authentic contributions from people um and that they have some skin in the game, we have

48. 23:41

    a staking collateral system. And it's a really exciting application of a web 3 primitive that rather than focusing on um just passive yield bearing, we're using staking as a way for the contributors to commit to quality uh meaningful contributions. And then after they uh do that, then they get their stake back, they get their rewards, and if they misbehave, then they get slashed. So it's a it's a way of getting like that real commitment out of people.

49. 24:15

    And then we have a skill tree KYC for being able to identify experts in different domains so that if a client comes to us in need of say dog experts, cat experts, um luxury good handbag experts, things like that. were able to run quick quizzes that for an expert in a field or um say a hobbyist, they could very easily answer these questions, but um if if you're not in the field, then um it's a little bit harder. So, we've

50. 24:45

    got that all set up and the community is active across all those touch points. So, that's kind of like how we're getting this done. Uh a little bit about traction and how things are going. uh we've got client revenue and that's a big differentiator with a lot of the other companies that are active in this web 3 uh AI uh data space. Um we have sold data sets to large enterprises and the community side growth is going great. We now have over 1.6 million data workers in our

51. 25:14

    community. Literally every continent and every country in the world um is participating in this which is essential for the kind of diversity that's required when you're seeking niche on demand data. And a little bit about the team behind this um myself uh some folks who have been around the near ecosystem may recognize my face. uh built um bunch of early DAO and NFT infrastructure on near. Before that I was at Google um

52. 25:44

    doing technical project management on material design and prior to that spent over 10 years in the advertising industry running a digital innovation lab under the Omnicom umbrella uh helping Fortune 500 companies do more interesting activations than just running banner ads. And my more technical co-founders, Dr. Steven Wong and Professor Suie um are actually the origin story of public AI. So they were running a AI research lab and ran into the problem of hey the data that we're

53. 26:16

    in need of is way too expensive and also when it comes back to us it's not of the quality that we need um to do our research work. So they decided to take a stab at a decentralized approach to getting that data. And after having some success, uh Steven moved to the Bay Area to find myself and Kenji to join and help uh turn this into a thriving healthy company. And Kenji uh has a deep history in more of the operational side

54. 26:50

    of things. He started Caro which is the number one largest used marketplace for cars in Southeast Asia and um on the business side has a a deep history on uh mergers and acquisitions and um he's our operational co-founder. So that's a little bit about the team, what inspired us to build this human layer of AI and why we're diving so deep into data today. So that's a little bit of background. Now, I want to get to some what does this look like? Can I participate today? Um, the answer is

55. 27:24

    yes. So, we've got uh campaigns running um all the time. And right now, we're in a break between the season 2 and season 3 of our uh audio recording for native language. So, what we're doing is um having people show up, record short sentences in their native tongue. And what this is being used for is real-time translation devices. If you remember the babbleish from Hitchhiker's Guide to the Galaxy that sits in your ear, um something like that. And this is a great earning opportunity for the community.

56. 27:57

    They can earn up to a dollar a day by uploading 140 samples, which takes under 30 minutes. And total, we're uh doing rewards over $500,000 for this. So, um, great thing to get started in now, even though we're about halfway through the seasons. Um, better better to get started now. And, um, this also brings up the point that as we're rewarding folks, we're doing a mixture of USDT and, uh, our points. So, as people come and participate in the system, it's not just airdrop farming.

57. 28:32

    They're actually getting like real rewards um, if they show up today. When you pop on to public AAI uh at beta.publicai.io, which is our data hub site, you'll have to set your mother tongue. Um, and this is something that can only be set once to make sure that people aren't hopping in between languages to do recordings.

58. 28:52

    Um, so make sure that you select the right one when you pop in there. And then as part of that uh verification process that I mentioned earlier, we have these quizzes running where after you select your native language, there will be a short quiz that ensures that you're actually a native. So super obvious questions if you are um and then after you've taken the quiz, you can go

59. 29:16

    to uploads and get started. The number one thing that our clients are looking for is that the uploads aren't generative AI and that people aren't using text to speech. um because those are things that they could do on their own, no problem. So, we do have a mechanical filter in place that looks for those things, but um still have people obviously trying to farm it whenever there's incentives. You got

60. 29:39

    farmers showing up. So, we're running our own AI models for doing that kind of detection and your uploads will be rejected if you do any gen AI or TTS uploads. And then as you're uploading um towards that 140 samples a day, we've got the gamified progress bar that shows you how you're doing and uh shows you what the next reward level is so that um you have motivation to keep on recording.

61. 30:04

    And then after that the voting process takes place and as I mentioned this is uh stake based voting. So to participate in the staking to prevent people from just jumping in and doing random clicks, uh we've got a stake mechanism in place and you get paid out after you reach consensus with your peers, whether that's a vote in the yes, this is good data direction or no, this is bad data direction.

62. 30:30

    And then after you've completed a few campaigns with us and understand how public AI works and want to take a a more active role in the community, we do have a upgrade tier system where you can become a guard in the in the community.

63. 30:46

    And there's also a judge uh system after that for people who are resolving disputes. So, um, if this is something that you're excited about, definitely dig in, give the current campaign a try. After you've seen how it all works, go ahead and apply for a guard role and take a more active role in the community. And then we also have a referral system live where um, our theory is that we've got seven degrees to any expert. So, um, this is for everybody in the world. uh as the human

64. 31:14

    layer of AI, we believe in capturing the value around the where where people fit into the AI work revolution and right now that the highest demand is in the data space and that relies on experts. So um no matter who you are, where you're living, uh go ahead and get signed up and then bring on your friends so that the community is full of experts in all different domains.

65. 31:42

    One more campaign that we're running right now is the hero yappers. So you can go um after taking a short quiz to learn about how public AI works. You can do a tweet about it, upload your tweet, and the more followers you have, the higher points you'll earn. And this is live right now, so go ahead and check it out. It's our our alternative to something like Kaido Yaps.

66. 32:06

    And as I had up on the screen earlier, we are really excited to partner with HOT on the first chain abstracted intense based marketplace launch and collection launch on the hotcraft.art uh platform. So database is minting on July 15th. It's a free mint and this is in celebration of hitting that milestone of over 1.5 million uh contributors. Uh even since then we have added another 100,000 folks. So we've got um 1.6 now.

67. 32:40

    And as part of that we wanted to be sure that this collection was for the community. So it's also under the viral public license which is similar to the creative commons that allows everybody to remix it, play with it, have fun, and be part of the community.

68. 32:58

    So that's it from me. Um, definitely give us a follow on X to keep up to date with all the latest campaigns and news. Um, join our Telegram community and you can also pop into Discord and uh, chat with our with all of our friends there.

69. 33:15

    So, thanks so much for the time today. Thank you so much for coming, Jordan. Keep up the good work and keep up keep on pushing public AI. Super excited to keep following all the progress you guys have going on. Next up, we got Owen from Proximity. Owen, my friend.

70. 33:37

    Oh, I'm on mute. Hey, how's it going? How's it going? Take it away, my friend, and let us know what you've been up to lately. Thank you. Thank you. Great updates from uh Public AI and Ilia there. So, yeah, uh Ilia briefly mentioned this before, but we've been working on the shade agent sandbox recently. Uh, and what this is, it's basically a collection of tools that allow you to get up and running with a shade agent as quickly as possible and with minimum required

71. 34:09

    knowledge uh, needed. So, just a bit of a refresher on what shade agents are if you've not heard of them. Uh, before like shade agents, web 3 agents fell in sort of one of two categories. The first were was that they were trustless and verifiable by using trusted execution environments. But if this TE went down, then the private keys and funds to the agent would be lost, which is obviously

72. 34:40

    not good. And then the other camp was uh the agents accounts were actually persistent. So they would remain if if uh wherever the agents running goes down. But to achieve this, the agents would have to be centralized. So shade agent framework allows uh developers to create crypto agents that are both trustless and verifiable by running in a TE and the private keys can't actually be lost or compromised. Uh so accounts are persistent.

73. 35:12

    And what does this sort of allow? What can shade agents actually do? So, first off, they're AI powered, so they can verifiably access LLMs, also running in TEES. They're multi-chain, so they can custody uh almost like any crypto asset and sign transactions on any chain using chain signatures.

74. 35:38

    And like I said, they're non-custodial, so not one entity controls the private keys of the agent. And this is using the the MPC signing um crosschain signatures we we all know. And yeah, like I've said before, they're they're verifiable. So, uh sort of like blockchain, the agent code is known and executes as you expect it would, and no one can actually Oh, I'm I just realized I'm not presenting

75. 36:08

    my slides. That was a a big mistake from me. Let me uh let me share my screen here. Here we're we're back. We're back. You just can you see that now? Could you put it up, please, Peter? There we are. Sorry about that. So, yeah, these are the slides. You didn't miss much. Um yeah, so the agent code is known and it executes as you expect and no one can actually tamper with that secure enclave running in the tea and they're

76. 36:42

    confidential like optionally so you can pass inputs and the state of the agent be can be kept private which allows some interesting use cases as we know blockchains aren't on private by default. uh they're very flexible so you can interact with off-chain data and APIs beyond what you could do within a blockchain like with a blockchain you can't just call an API but with this uh verifiable environment you can and they're low cost so you can run

77. 37:13

    intensive computation like run an LLM run some really complex uh data analysis inside this T without it costing you a load and without running into gas limits. So how do these work? We have these agents that sit in the tea. They have their own private key uh which is randomly generated based upon the keys like hardware entropy. So it's different each time it boots up and these can

78. 37:45

    access like LLMs and access APIs and data externally. And once uh once an agent boots up, what it does is registers in this agent contract. And this is a contract that actually lives on the near blockchain. And when it's registering, what it'll do is it'll submit a code hash and an attestation.

79. 38:08

    Uh the code hash basically says, hey, I'm running this code that you expect. Here it is. Uh and then it can match against the first code hash it was provided with to say are these two the same? Yes, they are. It's running the correct code. That's all good. Then we also have the atestation which says are you running in a TE? Yes, I am running.

80. 38:34

    If you're running in a TE and you're running the correct code, then I'm going to let you be uh a registered worker, an actual agent that's registered in the contract. And once it's registered, this agent can now access this sign transaction function which connects it to uh any blockchain. And the accounts associated with the agent are associated with the agent contract itself, not the

81. 39:04

    uh not the individual agents. And what power that gives it is that you can boot up another agent like the same agent that will have a completely different random private key. And as long as it's running in a T with the attistation and it's running the correct code provided by the code hash then it once again can verify and access these same accounts.

82. 39:28

    So this this uh system has a an aspect of livveness where if you if the agent goes down if the tea goes down you just boot up another one or you have loads running at the same time. So uh you can always access the service. So, to get on to the actual sandbox that we've been working on, uh, what is it? Like I say, it's a it's a suite of tools that really helps you get up and running with shade agents as quickly as

83. 40:00

    possible. And you really don't need that much knowledge on agents or even programming at all or specifically on N. We're really making it developer friendly from whatever background. And I challenge you, even if you're not a dev, to try and make something with this framework. And what it's made up of is a CLI, a command line tool that just deploys the agent for you, takes away

84. 40:28

    the complexities. We have a Docker API image that allows you to interact with the agent contract without having to know like near specific commands. We have we're using global contracts. So this agent contract is just deployed behind the scenes without you even knowing it and a library to to interact with all of this. So this means you can deploy an agent in just a single command and you don't need to know about about TE's you don't need to know about near

85. 41:00

    you just build your product and we abstract that all for you. And what we're trying to drive with this this sandbox is to make it as like language agnostic as possible. So you can create an agent in whatever language you want from whatever background.

86. 41:20

    So I'm going to attempt a bit of a quick live demo here just to show you what the sandbox actually looks like. Um so I think you can still see my screen. Yeah. So when you clone like the template repo, you'll be presented with this which has um your source file and this source file is is what's going to get uploaded to the agent. And in here we have we have roots. So the agent is basically just uh a standard server that's going to run inside this T. So we have a couple roots

87. 41:59

    here. This is an API route that will get the agent's account and its balance. We have one that will get like an Ethereum account and its balance derived from the agent contract because in this this template it's like a it's basically an oracle like a chain link oracle but it's the verifiability comes from the T instead of uh multiple nodes running the code. So we have an Ethereum address that pushes the price and then the magic happens in this transaction route and

88. 42:32

    what this route does it fetches the price of ETH from uh just some like Binance Coinbase APIs. It creates a payload and then it uses this function. I'm not sure if you can see but uh sign with agent and this pushes the the transaction payload to be signed by the agent. uh and we get a signature from that which then can then be broadcast to the target network in this case uh Ethereum.

89. 43:03

    And all you really need to do is edit these files to create your shade agent. And we also have uh a M file where you just fill out some details. So you'll need a near account. uh you can change this this prefix to your near account which your contract will be deployed to.

90. 43:24

    Uh you change this from like proxy to sandbox whether you're creating like locally you're you're developing locally or you actually want to put it on mainet. Um then we have an API code hash. You don't need to touch that. your app code hash. This is the code hash that's getting submitted to the agent contract and this is automatically filled out for you when you deploy. Um and then your docker tag which is just

91. 43:52

    your uh docker account details to host and build this image and then a fer key and we're using to host the tease. Uh so to deploy your agent once you're ready and you've built it, you can just run this command. It's shade agent CLI and this is going to uh build this image pretty quickly. It's going to push the image to docker.

92. 44:20

    So this is your like all your code bundled into a single file basically and it's going to deploy your your agent contract on testn net. You can switch it to mainet if you want. Uh you can see it's approving the code hashes in the agent here and then pushing it to fer cloud. So we get a link down here which we can open and see our deployment is actually booting up.

93. 44:54

    It will take just um just a minute here. to register up. But then we have this this endpoint and this is where our server is being hosted. Um and you can actually go ahead and interact with that inside your front end or you can interact with it from like the browser. So if I do slash API slash agent uh I think it's agent account. Uh still needs to boot up a little bit more. Once this is booted up, should be good.

94. 45:36

    Yeah, it'll give us the details uh of your agent account and then you can also do like transaction which will then obviously send that transaction to update the price of ETH. And we just have like a little a little front end here as well that you can go in to make it easier to interact with. So I'll just quickly host that front end.

95. 46:06

    Uh so once you host the front end, this is this is what you'll see. I might have put the URL in wrong there. Um I think without a slash. Yeah. So it'll show you like the current price of ETH and your balances and then you can click to set the price of ETH there.

96. 46:33

    And that is the the shade agent sandbox. So I was actually showing you the most recent version here and that will be uh deployed. It'll go live I think like end of this week or or next week. What we're really focusing on in this version is is modularity and uh flexibility so you can you can do more with the agent sandbox and then also getting it deployed uh a bit quicker. Uh and that's everything

97. 47:03

    from me. Thanks for having me. Thank you so much for coming, Owen. And I highly encourage everyone or anyone to go uh and play around with the shade agent sandbox. Um it's truly amazing framework and um I think we're only starting to scratch the surface of what's going to be built using this technology. Next up, we have Intel wallet. And um Slime is a beast. He's always building. He's always near. And he's here to talk a little bit about what uh what kind of world he's been working on. So everyone, give it up for Slime.

98. 47:48

    There he is. I'm doing great. real quick. Yep. Do you see this? Actually, I Yes, it's all it's all there. I'll do it like this. So, Interallet is um made by Interior. Let me do a quick introduction. My name is Slime and uh I'm here to talk about the release of interior wallet which was released this week and uh yeah it's not it's anial wallet.

99. 48:29

    It's open source and it's uh for a new blockchain. In here is um an infrastructure provider that uh has built um oracles developer APIs, indexing APIs, a lot of uh stuff for notifications in a telegram bot, a trading bot, um volume bot tool.

100. 48:54

     I don't even remember all the all the stuff we built. But u this time we're going to be talking about the inter wallet and I'm not here to say that in wallet is ready feature complete or on par with other wallets. I'm going to tell you about why you should switch from your existing wallet to in wallet. And the first reason is uh DEX aggregator.

101. 49:21

     Vex aggregator picks um the best price among all dexes and not only dexes but also launchpads liquid stacking providers and um other ways to swap tokens. I don't know why I made this as um random spin wheel. Uh it's not random.

102. 49:43

     Trust me. No, it's not. It's actually not random. So yeah. And more features that um I really want to talk about is um sub accounts. I'm going to show it here real quick. So I create a new account and uh let's say I want to create a for downhole instead of mirror I can select uh slam dragon.

103. 50:17

     and uh just create an account. And um yeah, also cute uh in intern or intern Betty backgrounds are included. So yeah, it's uh already here. I made a typo, but it doesn't matter. So yeah, by the way, you can choose the backgrounds here.

104. 50:43

     So yeah, next point is key rotation. It's also a feature of near that um not many wallets um recommend people to use but uh on interval it's quite easy to use. You just click terminate all sessions. Wait, let me get some mirror for this account.

105. 51:11

     It doesn't have to be 0.2 like most wallets require because um we understand that not not all people have this kind of money even of even so it's not really much but uh not everyone has um that much in crypto in year so it works just fine with 0.01 year you can even swap tokens with uh like 0.02 02 near and uh yeah so I click terminate all other sessions it will show me a new phrase which will be reapplied to my account. So for now this uh phrase is um starting with toilet

106. 51:54

     which is my favorite word from the dictionary and um I click here the new phrase will have mask. I just confirm and it will just replace my seat phrase without changing u account name without changing anything just a new seat phrase. By the way, if anyone wants to claim this account using this seat phrase uh you're welcome.

107. 52:24

     So yeah, we also have an explore tab which is um frequently updated which contains a lot of um depths on the new ecosystem like uh really a lot including like hotcraft and search swap which launched um really recently has um learning with a lot of articles And yeah, also it's a really developer friendly.

108. 52:59

     You can uh create um when you create a new test net wallet, for example, let's do this test net. It will preund this account with um 2.5 test net. So you don't have to find the wallet, find the faucet, which not not always works. You just have uh 2.5 near by the way.

109. 53:24

     Don't mind this number. It's just on test net. That's how much near cost in USDC on test net. It also has um some uh features for developers that makes that make UX improvements like when you unnecessarily split one transaction into two, it will uh display a warning that um UX might be improved for your app and um it's compatible with near CLI.

110. 53:53

     So yeah, just a lot of uh stuff. Of course, you can uh you can stake with a validator. You can st with liquid staking. You can see the APIs APIs. You can st with metapool and um it routes through metapool. So it's like built in into the swap feature. There are a lot of integrated base to swap tokens.

111. 54:25

     So you can choose um you can disable some but uh by default you just get the best rate. In the future we want to add a smart wallet which is feature that u was recently deployed on mainet here it's already here for test net accounts.

112. 54:50

     So if you click on settings account, you can enable smart wallet and um connect your Ethereum or Solana wallet. I will not do this uh right now because I don't have a Salana in this browser. But uh you can do this and then if you loging in with an existing account you can click Ethereum and uh sign a message and it will recover your test net account that you have linked to this account. In the future, we will have um

113. 55:24

     pass keys, Google, Apple accounts, and u some DeFi stuff, recurring payments. And uh yeah, by the way, it's free to deploy. On some blockchains, these accounts cost like $10, $20, $5, but this uh but here it's just free because of global contracts, which is a really amazing feature. Not sure if anyone uses it yet, but um Interwallet is um I'd say the first production if you can call test net a production app that use that.

114. 56:03

     If you use it on mobile phone, you can uh click add to home screen which uh will make it much faster and uh just uh feel much smoother to do this. It's called a PWA. It's uh it really speeds up everything if it's supported by the app and uh in in your wallet it's it is supported and uh in the future maybe this month we will have a native Android app and maybe sometime in the future and iOS app and desk desktop apps.

115. 56:39

     So yeah that's it. You can create your wallet now and um if you are interested you can follow up follow us on socials. We will have um detailed thread about Z wallet and its features um today on X and uh we'll have like a week of different threads about all the features. So make sure to follow us and uh create a new account using this QR code. Thanks.

116. 57:12

     Thank you so much for coming, Slime. I encourage anyone and everyone actually to go and test out the the Intel wallet. This guy knows what he's doing. Keep it up, Slime, and keep fishing. Really great to see. All right, next up we got Manuel from Silverstream. And Silverstream AI aims to provide the core infrastructure for reliable user aligned autonomous agents focusing on accessibility, privacy, and enterprisegrade automation.

117. 57:42

     Manuel, I'm just going to bring you on stage. Yeah. How are you doing, sir? Good, good. Uh, am I sharing my screen? Are you seeing one one second? I'll just remove mine and I'll add yours to the stage. Now you are take it away sir.

118. 57:59

     Nice. Uh oops wrong button. Yeah like as Ne was said we do web agents. Web agents are agents that interact with the web. They click and type like people do and we really focus on the reliability part of making sure they do what they need to do every time.

119. 58:17

     Um, I was really happy to see efforts like public to try to get more people into the system into into AI labeling and data collection because that's that's something we need to work on. So, it's great to get people in the ecosystem working on what's important.

120. 58:35

     So, let me talk about Silverstream. Basically, one of our agents start interacting with a query. So, a user asks something to the agent. They request a new MacBook Pro and in the case of the near app then the user talks to the app. The app goes to the back end. The back end goes to our agent and inside our agent which is in this trusted execution environment encrypted box. We have the agent itself doing the orchestration on everything that needs to be done and the agent as a browser

121. 59:08

     with whom that that can be used to interact with the web. So it will look at a web page, it will interact with it, clicking buttons and typing whatever needs to be typed to get the MacBook Pro and when it's done it goes back to the user and say thank you.

122. 59:25

     This is the the the laptop was sent and so on. Right? So it's a closed loop of user interaction, agent interaction and internet interaction. We send back screenshots and conversation but it's mostly between the agent and the internet. And so this is a useful way for your when you build your own agent to interface with the open web. Not everything has an API. Not every API can do everything on the platform. With web agents, we can now open up all these closed silos of data of goods and

123. 1:00:00

     services. Let me show you how our agent sees the web. So this is a simple homepage. On the left you see the image of the page itself. On the right you see how the agent sees the page. So we mark all the interactable elements. We look at the DOM to see what's clickable, what's typable and so on. And we give it a little bit a little number. Then we look at the HTML code and we turn that HTML into this markdown representation of 735. It's a button.

124. 1:00:36

     It's called toggle menu. And if you want to click it, these are the coordinates. We add as much information as we can to the context of the of the agent. And that's where we spend most of our our work. We also look at extra stuff like network request. This is useful for the agent to understand what it has done in the previous interaction, console logs and whatever we can think of. So our agents understand a web page very well

125. 1:01:02

     and and then they take actions on it. So in this example um you see the user is asking about flights from Montreal to New York. The agent goes on the internet goes on Google flights with the browser. So the first action will be page go to Google flights or flights.google.com google.com and then sees the HTML the X3 which is the description I told you before and the screen show with whatever is clickable and then the agent sends code to the page. So it can either click on the page

126. 1:01:36

     by coordinates or by those numbers we defined before or can write his own playright code and this action observation loop continues until the task is done. Now let me before going to that let me share. So this is the agent as as it works on a simple task.

127. 1:01:58

     Can you see the video? I Okay, nice. Uh, so this is our interface just for sake of uh explanability, but then it's also on the near uh year for example. My colleague ah let me go back. My colleague asked, can you go on the developer instance of service now log an incident uh reported to par peril and and whatever needs to be done. So Sylvester our agent takes it on thinks about it a little bit and you can see what's happening on the back.

128. 1:02:38

     It's going on the service now page. There's a thinking process to communicate with the user. So we we keep some uh read only update for the user to be to be updated about what's going on. But in the back it's clicking and typing around up until it feels the incident report.

129. 1:02:59

     Yes, there we go. Service Now does not offer an API, but now we can interact with it through an agent programmatically. And these agents can be chained on the near app. You can call other agents like ours to do these tasks that are outside of what a usual API system can do. An example people asked us is for getting APIs to get pricing of coins on um on systems you cannot integrate yet. So it's that

130. 1:03:33

     we deal with that long tale of systems that you cannot integrate today. Maybe you will be able to integrate in the future. So I'm very happy to have conversations with people and developers in the ecosystem to to integrate with to have your system integrated with agents.

131. 1:03:50

     Our agent is meant to be open source. We didn't open source it yet because it's not as reliable and stable as we want, but it's going to be. So I'm very happy to share you know source with people we work with um because that's where we're going towards.

132. 1:04:07

     Now the last point uh we are short on time. We just had a breakthrough performance. This has been really the effect of working with people in the ecosystem. People have asked for many things and every time there is an issue with the agent there is an improvement in the capabilities. Right? So over the last year we went from 63% uh top five per performance to 94% top five per performance. 63 was the best alternative before us and now we really

133. 1:04:39

     jumped almost to the end like at 94 the last 5% is is less way less meaningful than this first jump. And these are tasks like like the one you can see on the right. It's a third party uh benchmark by service now themselves. So create a filter create these are common task people do on their platform.

134. 1:05:01

     They create a benchmark for it for people to work on. If you want to see more of these um this bench these tasks this is the link. It's pretty easy to copy so you can just look it up. Uh in general the near community has been very helpful with feedback and collaborations. We work with some people in the community already and we are eager especially I'm eager to help people in this community because you guys think of things in the right way openness interoperability uh privacy and encryption those are the

135. 1:05:33

     things are the future of the web and so I want to be as close as close as possible. Thank you. Thank you. Thank you very much for coming, Manuel. And um I encourage anyone who wants to look into technology, reach out to Manuel. He's here to work with you guys.

136. 1:05:54

     And incredible to hear how um supportive the community has been. Really um warms my heart. And next up, we got Caesar from IQI. Cesar, I'm just uh you're on. Take it away, my friend. Um hi uh thank you so much for having us. Uh great work everyone in the previous presentations. Um I am uh Cisar Rodriguez the CTO of IQ. Uh in IQ we have the mission of connect AI and blockchain and allow developers to tokenize their agents. Um for that we

137. 1:06:33

     have created a ecosystem uh governed by IQ. Um on the screen right now I think you can see uh our landing page for the ATP which is the launchpad that we have uh for tokenized agents. Um I will go through it a little bit. Um here you can see uh or perhaps I need to change.

138. 1:07:02

     I think I have to share another hold on. Um yeah here you can see our ATP like the launchpad that we have for the agents. Um we have a different agents that we have built like Sophia or Adam. Uh Sophia is an agent that we built using our framework. Uh that allows uh create wikis automatically in IQ.Wiki which is another platform that we have.

139. 1:07:40

     Um here you can see uh for example all the logs that happen uh when the agent does a transaction on chain to create wikis and so on and all this is built on ADK which is another part of our ecosystem which is our framework uh to build agents is a very modular framework that follows the standards and allows stuff like memory control multilms multi- aents um telemetry everything that you need to go uh live uh with agents. Um that's coming from uh the

140. 1:08:16

     last two years that we've been building all these different agents and that's the reason why we ended up building our own framework because we were facing some issues over these years and we ended up with the framework. Um we also as part of this framework we have built um a lot of MCPS uh which is something that I want to demo today. Um the MCPS allow uh to connect our framework with the blockchain. Uh we are building MCPS that are specialized on blockchains. Um

141. 1:08:48

     the one that I want to demo today is using the near intents. Um we thought that it was going to be very beneficial uh to build MCP for intense. So because MCPS are like the standard for the AI uh to connect outside. Um so yeah, we create this one. Um and I think let me share a little bit of this.

142. 1:09:18

     Um um yeah um so here I think um you can see uh how it works. This is using our framework and the MCP intent. Uh in this case uh we are going to swap uh one year uh to USDT. Um here um we have in this case we we asked the the agent ask for permission. Uh in other cases this is a proof of concept. In other cases we just automatically do everything. Um but here you can see um how uh you have the near now instantly you get the USD. Um the

143. 1:10:07

     cool thing about this as as I mentioned before is that this using the standard MCP. Uh so even if you don't want to use our framework, you could use any other framework or you could even connect it to cloud or probably any other framework to build agents. Um this is one of the two a MCPS that we have for we have another one that is specialized on connecting to to the blockchain itself. um which um perhaps we can demo

144. 1:10:39

     in in the next town hall. Um we have in our road map we have a bunch of cool stuff to to work with near um that I think we want to release uh or showcase in the next town hall even one connected with robotics which is something that we are building in our R&D department. Um but I would like to show it I think in next month. Um I guess that's all. Um something I will recommend for anyone who wants to get involved go to adk.iqai.com

145. 1:11:17

     or iqai.com and read more about us. Um we are also planning to run some hackathons in the upcoming weeks. Um that's something that uh if there are some developers there that are interested in building using our framework or using the MCP for near or improve the MCP for near intents uh they are very welcome. Um so yeah I guess that's all from us. Um thank you for much having us.

146. 1:11:49

     Thank you so much for coming. Have a great night. All right, next up we have Stephanie from the near events team here to tell us everything that's been going on and especially all the dope stuff they've been doing lately. Stephanie, let me know when to switch the slides.

147. 1:12:09

     Awesome. Thanks, Peter. Uh, go to the first slide, please. Awesome. So, GMGM everyone. Uh, Ilia alluded to this a little bit earlier on the call, but I'm going to walk you through all our of our in real life activations uh from permission links to ECC and pretty much how that all funneled into like one cohesive narrative which was um part of our fork that campaign. And as you recall, fork that wasn't just a campaign. was more of

148. 1:12:36

     a rally and crying. A rally and cry to make sure that builders and rebels and every everyone who kind of looked at the default settings that society offers um and says, "Hey, that's not right. We need to build something new, a new paradigm." Um, next slide.

149. 1:12:51

     So, I'm going to go over for permission list, the goals, the engagement and traction, our speaker highlights, how the stunt boxes were received, some survey results, and then some feedback from the the community and folks that attended our events. Next slide, please.

150. 1:13:06

     There we go. So, the goals, we wanted a three or more branded touch points um that really hit home for people that when we were when we were in these in real life activations, people understood that it was all part of the fork that campaign and you'll see that from our branding to the swag we gave out, but also to the folks we decided to partner with. Um so, as I mentioned, two key key take uh takeaways. Fork that and near is the blockchain for AI. Next slide.

151. 1:13:36

     Next slide, please. Um, so we had a slew of events taking place um, in New York City during permissionless. We kicked things off with the near A100X investor dinner. Big shout outs to Ardum on the uh, capitals team that helped us put that together. Uh, then the next day we went over to the Thai and the near institutional breakfast. Big shout outs to Mark on that one. Uh we ended the day

152. 1:13:59

     with a Bidd uh Biddle and fork fat founder dinner which was pretty well attended. You'll see some pictures in a few seconds. And then we also had some case studies. So shout out to Jared Thompson on the foundation team that uh helped set that out. I'm going to be also sharing uh some content from that and he'll shortly be sharing videos in the next week or so. Uh and then on the 25th we had a forked happy hour that took place right where permissionless

153. 1:14:24

     was uh being held in industry city. And then as Ilia mentioned earlier, we had the open source AI summit which we co-hosted uh with Erica. And then finally, last but not least, we had the near eco dinner which was a nice place to see folks from the ecosystem in New York City and those that were uh in town visiting for permissionless. Next slide.

154. 1:14:44

     As I mentioned, the institutional breakfast was went pretty well. Uh Ilia had a gave a brief little talk uh talking about near what Near AI is up to. It was pretty wellreceived from what I understand from the part uh from our partner the Thai. Um it was held in the city at Little Maven on the 24th as I mentioned. Um and we were being very selective. We weren't really blasting this event out because we didn't want to

155. 1:15:06

     be inundated with a lot of a lot of people registering and then kind of bothering us as to hey why am I not accepted etc etc. So we were being very intentional with our comms on this and be and not really blasting it out. Um 100 people registered 70 approved and 40 checked in. Pretty good odds I'd say.

156. 1:15:22

     Uh, next slide. Uh, as I mentioned earlier, we also had a bidd and forked that uh, founder dinner later that evening at rule of thirds. If anyone has ever attend been to that restaurant in Williamsburg, it's really really cute. It's also bomb food.

157. 1:15:38

     Everyone enjoyed it. Um, Erica and um, Ilia were the hosts which was awesome. It was also a nice place to since the that campaign was revealed on the 23rd. It was also a nice place to give away our some of our stunt boxes. So, you guys may recognize some faces in the in that photo like Alton, uh, Dean, and a few others who were really happy with the gift. Dean actually arrived first,

158. 1:16:00

     and I have a funny photo that he sent me where he wanted to take all the all the treasures as he said they he he would have walked away with them, but then people slowly uh started piling in. But Dean's always a great person and a and a friend of the ecosystem. Uh, just for some stat purposes, 20 folks were invited and 14 showed up. So, that's a great success rate. Um both Ilia and Erica send out those invites. So thank you so much um for supporting our events. Next slide.

159. 1:16:27

     Then we had a fork that happy hour. So um as I mentioned this was a co right next to where permissionless was being held in industry city. Uh we partnered up actually with ACJR as well as cracked labs. ACJR for those not familiar um is the association for crypto uh research and journalist. Um so they brought media to the room which was awesome. got to connect with a few folks there made and made good connections. Uh and then Crack

160. 1:16:53

     Labs was actually the partner for the hackathon taking place at permissionless. So they were able to funnel that and push that to the dev community. So it was a nice turnout. Uh as I mention checked in again pretty good show rate all things considering. Uh next slide please.

161. 1:17:14

     Open source AI summit. So, as I mentioned earlier, we were a co-host, which means we um helped with some of the curation of the agenda, uh but mostly added support in marketing purposes. So, we actually made the uh stage banner that you see back there, which is why it's it's big shout outs to Charlie on that, which is why we have a big nice splash of presented by Near. Uh some of the notable speakers, Ilia, Tram, Eric, See, Casey, a slew of many

162. 1:17:40

     more, but I couldn't add them all on this slide. uh trying to be respectful of your time, but overall pretty good success. Um the the Block Works team was quite happy with the caliber of speaker speakers speakers that we were able to court. Um and hopefully we could get this opportunity again uh next time around. Next slide.

163. 1:17:58

     The near uh eco dinner was at the beautiful Laser Wolf. Again, this is one of those places in Williamsburg that if you haven't been, take the time to go next time you're in New York or if you live in New York, make a reservation.

164. 1:18:10

     wonderful view, delicious food. It was a great way to spend time with folks in the ecosystem, connect. Um, I got to see Benji, which I haven't seen in a while. Um, and he gave some alpha that he will be back living in this side of the world very soon. So, very excited for Benji if he's if he's on this call. I'm not sure.

165. 1:18:26

     Um, but it was a great time. Big shout outs to Maggie for organizing that. Thank you so much. Next slide. Content, content, content. As I mentioned, Jared was able to do some founder videos. So, he did a really great job. He'll be releasing more of the full videos soon if he hasn't already. I might have missed that update. I was in ECC, so my Slack messages were a bit messy, but um he did

166. 1:18:49

     a great job. Shared some of the capturing mo some of the moments he captured uh during the sessions, and he said it it was probably one of the best um recording capturing of founders so far. Um especially in as the fork that campaign came out, people were really able to share their message, share what their company did, uh their project does and then tie it back into the fork that moment. So really nice timing in our end. Next slide.

167. 1:19:15

     Uh next slide, speaker highlights. Um as I mentioned, uh we had a lot of in real life activations. We also had speaking opportunities and Iliot was really everywhere all at once. He spoke at the Akos accelerate um event. Uh and then he then he also spoke on a panel at the permissionless conference. Uh next slide.

168. 1:19:39

     There we go. And then he had two speaking opportunities at the open source AI summit. He gave a keynote solo and then he also was in a uh I guess fireside debate of sorts with uh Michael Casey and Shri Ram from Ian Layer. So really he was he was doing a lot the whole week but so shout outs to Ilia for uh taking the team uh taking the near team home. Uh next slide stunt boxes.

169. 1:20:04

     Let's Yeah. So as I mentioned we gave them out during the dinner. Uh it was really well received. People loved the messaging, loved the box, loved the creativity behind it. The fork that button was actually really cool. Here we have Tommy um from Deli uh giving us a shout out for the creativity for for the messaging. And also for those that aren't aware, we actually gave away um some near via intense. So people got to

170. 1:20:27

     experience the technology and how frictionless near intense is all in one go. Um and it and we intentionally made designed the box in this manner. So it was light, easy, and just that button and with the goal and the hope that people would uh tweet about it and that's exactly what happened. So that was a really awesome moment. Uh next slide.

171. 1:20:48

     Next slide. Survey results. So, um, we wish we could have gotten more recorded responses. I did try my best on site and to follow up postevent. Uh, permissionless is an event we don't normally go to and so we don't have as big of a like a near eco uh, Telegram group chat community, so it proved a little bit harder. Um, but overall people really liked it. both anyone who was a partner or media or a hacker or

172. 1:21:12

     just a regular attendee really seemed to enjoy the food um the ambiance how close it was and just the lack nature of it. We also gave away some swag that we had left over um from some of our other activations. So people folks were really gung-ho about the overall experience.

173. 1:21:28

     Next slide. Um survey results just kind of what we always ask like what was good about the event, what was bad, etc. Uh don't quite know how they got that stuff as the best in. have to figure out what happens to that survey, but that's not part of the categories. Uh networking opportunities was awesome. Venue, location, and setup, the food, and then uh programming just like how we structured it. People really seem to enjoy as I mentioned some just like casual description. There's three

174. 1:21:54

     ways to describe near good vibes, open forward thinking, agentic, interop, intentful. Um folks are really resonate resonating with our messaging. Next, yeah, next slide. How can we improve more dev focus activations with content or workshops?

175. 1:22:09

     So, one of the things that we did get feedback like and this was in in real life but also in the survey was that near does a good job of bringing devs in the room and and they like the vibes. Um, but some we've compared to last year we've done a little bit less content and workshop things for some of our activations. So, that's something that some founders and some folks just want to see again. So, we're trying to think of how the next the next event we incorporate more of that um either workshops or content or fire sites, etc.

176. 1:22:37

     Uh but trying to keep in mind that most of the time when we do put that on, people are still just yapping in the back. And so, how do we strike that that balance and and we're working on it. And then what we can improve, um you know, we did have other options besides lobster rolls, but I guess this person really was upset that they were available. But again, if you have the allergy, my two cents is don't eat it.

177. 1:22:59

     They were separate enough that you wouldn't it would there would be no contamination. But overall, food was great, drinks were great, and the vibes were awesome. The venue did a really great job of uh helping us out through the whole process. So, if we end up going to if permissionless takes place in Brooklyn again in the same area, uh we would love to go back there. U next slide. Yeah. So, that wraps up permissionless.

178. 1:23:19

     Thanks. Um now, we're going to segue into ETC because we went backtoback events. So, uh next slide. Next slide. Similar breakdown. We're going to go through the goals, engagement, speaker highlights, the booth, because we actually had a booth at this um um at this conference, stunt boxes, some survey results, and what what worked well and where we can improve. Next slide.

179. 1:23:41

     Similar to the last event, but we just raised the stakes a little bit. We're going to hoping five branded touch points. Uh fork that is the rallying cry for near get folks to recognize near as the blockchain for AI. And then we added a third since this was a more intent focused conference uh where we want we want people to understand what near intense is um why it's working and hopefully how they can integrate it into whatever they're building. Uh next slide

180. 1:24:06

     next slide. Uh we had a lot of events taking place this time. Uh Oz City agents day agents unleash the house of stake delegate salon. uh the intense happy hour and then finally the pantra in unit 410 happy hour. Next slide. Uh a city big shout out to Shai who I think was in France for probably two weeks at the by the time he left. Um really held it down for for the foundation. He hosted an intro to near session. He did an intense workshop. He did a chain six workshop. Also, big shout outs to Maria and team who helped

181. 1:24:41

     with um putting together bounties and and helping that get going. Um she shared this where um the final I guess activate or the final numbers was six great results or six great submissions, excuse me, for the near bounty. Um, and from what I heard, at least from what AVB told me, I haven't fully confirmed this, but one of the winners was someone who really had, it was, I think it was like a working mom who's never really had experience in building, but just

182. 1:25:06

     found the tech very easily easy and and accessible and ended up doing something really cool. Uh, I know there's pictures floating around in the chat. So, I'll take a look and maybe add them to the at the end. But, if you're curious about what the submissions look like, feel free to take a look at the um in that link and then just filter by bounty.

183. 1:25:22

     Next slide. near eco kickback. Big shout outs to Conrad and team as well as Conrad's wife who was really really an amazing host. Um so this was a bit impromptu was not expected. Um I actually landed the day that they everyone was just kind of like hey let's just do a kickback at Conrads.

184. 1:25:43

     Um it it ended up being really really fun. Um and I think the impromptu aspect of it actually lended itself to be more relaxed. Um, so it was at the market success villa because they actually had an off-site beforehand. It was just near eco and friends. And as I mentioned, this is probably one of the most recent big near eco and and foundation photos.

185. 1:26:02

     So big shout outs to them uh for putting this together last minute and just, you know, and just inviting us to hang out and because that's kind of aside from building really cool stuff, it's also building really great connections with awesome people. So thank you. Uh, next slide.

186. 1:26:15

     House of Stakes. Big shout outs to Lane and Claus and everyone else who participated. make sure this took place and happened and organized it. Um, so it was held at the same hotel as the happy, excuse me, at the same venue as the happy hour. It was from 9:00 to 5:00. So they were just hitting they hit the ground running actually when I showed up to and shout outs to Olga as well for helping me prep the room when I when we showed up to just make sure everything was in order and and help prepare

187. 1:26:39

     anything like the swag and whatnot. Um Klaus actually rolled up with a huge piece of parchment for workshop stuff which I haven't seen that kind of I guess uh workshop style in a while um since organizing and more community stuff. So it was really really fun um just from an attendee stats perspective 25 in invited 20 showed including folks from the foundation both from legal and finance to give any advice to um to the

188. 1:27:03

     house of state uh crew delegate crew uh just to make sure that the project gets off the ground and governance works. Next slide. the near intent uh intent powered happy hour. That's a mouthful. Um yeah, we held it at the same hotel um at from 4:30 to 7. Just some attendee stats. 960 registered, 240 accepted, 140 checked in. Really great numbers. Uh one thing to note about ETC, there were a lot less events this year, which I think helped with getting the numbers where we where they are right now. Folks really seemed to enjoy the the view. was beautiful.

189. 1:27:39

     Especially if you stayed longer, like right around sunset, you got that super awesome uh you know, purple orange view from from the rooftop. Drinks were good, food was good, and as always, the company was the best. So, thanks so much for everyone that showed up. Um I know that we uh we partnered with Aurora, so big shout out to Aurora for sponsoring the event as well as Yap Global for sponsoring the event. They Aurora brought in the devs. Yap Global brought in the media. So, thanks so much. Next slide.

190. 1:28:08

     Crypto on the coast. Uh big shout outs to Mark for helping this uh put this together. Pant pantering unit 410 were our co-hosts. We had a pretty good showing. It was right on the beach. They did actually actually have to update that because they did it from 6:30 to 9, which was the perfect timing for those that were at ETCC. It was quite hot the whole week. So 6 is finally when the sun was kind of starting to break and and

191. 1:28:30

     not and not slay us too much. Um but yeah, attendee stats 450 registered, 200 accepted, 100 checked in, really good vibes, really good um mostly I would say institutional investor type of folks. Um but for the founders that were able to come in and make it, I think they made some great connections and hopefully that leads to some raises. Next slide.

192. 1:28:52

     Speaker highlights. We had a lot of folks speaking, Ilia, Lane, uh V from across the ecosystem. So, I'll just start off with one of the first speaking opportunities for Ilia was actually a debate with Justin Drake. Um, and I I actually didn't get to see it because I was running around doing other stuff.

193. 1:29:08

     So, I don't know who won, but just let's say we're a little biased and just say Ilia won. Um, but it would looked like a really good turnout. If you guys are interested in uh viewing the actual debate, it's available online. Just look up the name of the of the session and then just put ETC and it'll pop right up. Um, V over from Proximity spoke at L2CON. Um, and actually his shirt, which actually just says Nirvana, but people want to do a Nirvana shirt, which I'm going to work on it. Marketing has heard

194. 1:29:33

     the cries of the people, and we will try to get a a Nirvana shirt for our next event. Um, was really popular. He did a great job with this presentation. So, shout outs to V. Next slide. As I mentioned, we're really everywhere and Lane was no exception. He had a debate of his own. Again, I was running around, didn't get a chance to see it, but again, similar bias. Since he's with Near, I'm team Near and I hope he won.

195. 1:29:56

     And if not, you can check it out online and see where and give your own opinion on who won. Um, and then he did a solo uh uh presentation called Hail to the Chief. Really clever words there. Um, play on words there. Um, did a really great job for the again available online if you guys are interested in viewing it. Next slide.

196. 1:30:15

     And as always, I try to spread the love with all speaking opportunities. Shout outs to Jay from Nei for moderating this session at the Hack Seasons conference. For for those of you that haven't had the chance to see Jay moderate or even speak on a panel, you you're missing out. Like you really have to go because he he's spicy. He does a lot of hot takes. Uh and and it kind of throws a throws a curveball to the panelists every single time. Uh so he did a great job and and the actually the organizers

197. 1:30:39

     were like, "Please, we want him back." So next event definitely going to put him in a moderator role again if if if possible. Uh P Pierre actually did um uh a solo of his own at the encryption day event. He did a really great job. He he actually showed AVB and myself the presentation beforehand just to kind of get some feedback. And I just want to say that big shout outs for Pierre when I started placing him in speaking opportunities. He's really grown on his

198. 1:31:05

     own. So I'm really really proud to see that he's just continues to excel um and growing in that aspect of of becoming more so assured uh during his presentations. Next slide. ETC. Um, Ilia had a solo for the I was actually able to attend that one. Um, so how blockchains can make AI user owned.

199. 1:31:25

     Really amazing presentation. Was pretty well attended. He had a one or two questions I think post presentation. Um, and really and I think it he did a great job answering them although I think he was a a little bit cut off towards the end but you know can't those are the organizers. Can't do anything about that. But shout outs to Ilia for getting that done. Um, we actually had a second opportunity at the Hack Seasons conference. Lane was a panelist for building AI that lets the chain think

200. 1:31:48

     for itself. Um, I think that's a that's a I'm not going to lie, Lane, that's a little bit scary, but we'll see. We'll see where it ends up. I haven't had a chance to review the content for that one, but I'll take a look post this event. Next slide.

201. 1:32:01

     Um, and I wasn't able to find a picture uh for this one, but Lane actually participated in this game theory aligning a sentence for success um panel. It was a little bit last minute, but I'm sure you did a great job. And I'll actually get a more in-depth view once I uh view the footage. And then um Ilia also participated in the fireside chat um at ETH Pragma. We'd have a better picture, but unfortunately there was a fire drill like right towards the

202. 1:32:27

     end. So I think this is the best picture we got from Erica. So big shout outs for taking it. Um and then yeah, I think he he gave too much alpha and the the roof caught on fire. I'm not quite sure, but um it is what it is. But thank you uh for those where that participated. Next slide.

203. 1:32:42

     And then the booth. Next slide. Uh we had a fun time at the booth. Uh we haven't had a booth experience in a little bit. Um but we we came back really strong for ETCC. Big shout outs to Jay. Uh this man can sell water to a well. He really helped us move the swag, which would that was part of the goal like getting people to sign up via the QR code, but also getting people to be excited about our swag and talking about Near and Near Intense did a really great

204. 1:33:07

     job. Um, Ilia actually the man myth legend himself showed up to the booth um and getting showed some support and love. Uh, nice picture there with Esther Davin and AVB. And then uh big shout outs to Aurora for helping support the booth. Uh, Nick uh, Avan and Armen came and showed love and also helped us just staff the booth uh, when some folks have to have speaking opportunities or myself was running around doing other things.

205. 1:33:32

     So, thank you so much for the support. Next slide. Next slide. stunt boxes. So, as we mentioned, we gave some of them out during dinner at Permissionless, but we brought some to France and we gave a few boxes to uh Samantha uh from Gap Global who gave them to journalists. And so, we actually was able were able to put one in the hands of um McKenzie who is a cryp I guess finance crypto journalist over at MSNBC, a friend of the family

206. 1:33:59

     hopefully at this point. And she really loved the box, loved the happy hour, loved the views. So, big shout outs to Samantha for helping distribute those to the journalists that attended. Um, and then Erica and Ilia, big shout outs to both of you for uh handing out the rest of the boxes during some of the uh dinners and and private invitations you guys had. As you can see, they handed one to Sandeep and he really really loved it. Tweeted about it, really resonated with the messaging and then,

207. 1:34:23

     you know, was playing the tech which is the whole point. Um, so these boxes have been wellreceived and the momentum continues. So, we're really proud of that fact. Next slide, please. Next slide. Survey results. Uh, pretty good participation. We had 17 this time around and pretty mixed I would say from attendee to partners to media hacker staff etc. Um, pretty over overall good score. Um, I think pretty happy. There's

208. 1:34:50

     always room for improvement. So that's where we're going to segue to the next slide. So again, networking opportunity was great. Veny was great. Food and beverage was great. Uh and then from the three words it was just kind of like subtle sub subtle irrelevant focus um vibes were smart the sun people really love that that view and and there was AC at our event which not all events had definitely miss and then just the great and relevant relevant activation. I think we we timed it well we picked a

209. 1:35:19

     good location and people really enjoyed the fact that we were considerate that it's really quite hot in the French Raer that time of the year. Next slide. Next slide. uh one of the they some I mean yeah some of the feedback was a bigger venue and while we would love to host a bigger venue um it's also trying to keep it be mindful that we are trying to curate a specific experience and so while we can

210. 1:35:46

     ho we can get a venue that fits 200 people but then it's like are you going to be able to chat with 200 people are you going to be able to actually connect probably not so we're trying to keep it intimate so while we do uh take all feedback seriously we are trying to keep in mind of the of the of the vibe that we're going for. Uh, one I will say closer to the main venue. It wasn't that far. It was about a 12 15 minute walk, but I get that in the heat it was it probably felt a lot further. Um, but you

211. 1:36:11

     know, there's trade-offs. The venue is closer to the the the main venue didn't have as nice views. So, we'll take that into account and consider it if we end up going to ETC next year, but because it will also be in the it will once again be in the French Veriera. But, just keep that in mind. There will be trade-offs if we go one way or another.

212. 1:36:28

     Uh what can we improve? Some of the some folks thought the food was too light and only one bathroom and as someone wrote much um two of those things I cannot control. One I can so for the food too light we will take that into mind and try to maybe gives heavier more portions or a heavier um type of cuisine so that you know while you're drinking you're not you're not too intoxicated. Next slide.

213. 1:36:52

     Ah that's actually it. Well thanks everyone. That's the wrap-up for both permissions and ECC and how it all tied in to the fork that campaign. Appreciate your time. Over to you, Peter. Thank you so much, Stephanie, and uh great to see all these great activations all over the world. I was lucky enough to attend ETC and I really enjoyed the vibes and everything NIA had going on. Keep up the good work, NIA marketing.

214. 1:37:21

     This is what it's going to take us to the next level. Next up, we got Thomas from Omni Lane. Thomas, hello Peter. How are you? How are you doing? I'm very good. What about you? I'm doing well, thank you. You feel free to share your screen. Yeah, for sure.

215. 1:37:49

     Hello everybody. My name is Thomas Masello, ambassador in Buenosides, Argentina. So while I was listening to Stephanie, I was thinking, when are you going to do a near event here in Buenosides, Stephanie, we should talk today. I am the product leader of Omni Lane and I want to talk about some stuff about Omni. Okay. So the agenda is uh short. First of all talk a little bit about what's on then our current status.

216. 1:38:25

     Then our main learnings the last months our role in the new ecosystem. What we need from the community of near and last but not least a quick demo of how omni works. Okay so h in a nutshell what's omnilane? Omniame is a swap that is built using a near intense and multi- signatures infrastructure.

217. 1:38:54

     Uh we do atomic swaps in a fast and secure way. We can do crosschain swaps between lots of blockchains and tokens. And uh for example with omniame in quick uh steps you can do for example uh exchange of near in near blockchain for USDT in base blockchain or maybe you can do even uh USDT from base to bitcoin just in a couple of h clicks. That's the main idea of omni lane and h it's quite similar to h new intense but we want to make h the user experience even better.

218. 1:39:35

     Okay. What's the current status right now of omni lane and here I want to be open and very honest to everybody. Uh we have heard some erh bad comments about Omniane because uh our community building and our go to market h plan has not been uh how we plan it. So I want to talk to you with uh being the most open possible and tell you what's about omin right now. Okay. Erh what's good about

219. 1:40:11

     Omni Lane is that we launched our product uh some months ago. Uh the product has always been uh online. It was never turned off and uh we have done several swaps. What's been wrong? H the things that I told you before we created a go to market plan it doesn't go the way we planned. H so we are fixing it. H that's why I am here uh putting my face in front of all of you because we want to be alive and we want to continue uh working building our community. Okay.

220. 1:40:49

     What are the key learnings we have the past months besides what I talked to you that community building is not optional and actually it is critical. We also h be learning about h our user experience and how we can improve it. So that's why in this moment erh for everyone that want to that want to use omni lane I invite you to go to our page and try to do a swap. It is actually easy. However, we know that there are some swaps that are

221. 1:41:23

     more validated that its user experience is easier and we want to reach that you know that's why right now we are doing some feedback chats feedback calls with Omni users just to understand uh what are the best things about Omniane what are the blockers that they find in Omniane and how we can improve this kind of stuff just to make our product better you know. Erh then I want to talk a

222. 1:41:54

     little bit about what is the role in Omni in your ecosystem and why I think it is important. Okay. Erh everyone here know that new ecosystem is a blockchain that is growing a lot and it's a defy ecosystem is expanding faster you know uh with technologies as near intense and multi-arty signatures being like the the main technologies developed in this de ecosystems omni lane is a perfect use case of how

223. 1:42:29

     you can create for example a um native tab of near use this type of infrastructure as near intents or multi-party signatures and create a DAP that is useful that can create a you know um different goto market plans or strategies to make community and make grow this ecosystem. Erh last but not least uh before our demo I want to uh talk to you what we are searching for. H we want uh people to go

224. 1:43:09

     to our dab to try our dab to make a swap to do whatever kind of swap they wanted to do h and give us row and honest feedback. Okay. Why? Because how I told you some minutes ago h we are uh trying to make omnib. We want to make its user experience better and we want people to try it and to say to us hey I like this or mate this is a for example because that's uh how this h kind of products grow h because of its

225. 1:43:45

     community. Okay. Uh here uh you can see some uh links to our socials omnil labs in x uh omni in telegram and omnilane.x y and sit in our our landing. Okay. Now I will go to our landing page and I will try to do a swap in just a couple of clicks so you can see it. Erh here you can connect uh either your near wallet or your EVM wallet. In this case, I'm going to connect my Meteor wallet.

226. 1:44:24

     Okay, here you can see that I have I think like five near. Yeah, I have five near almost six. I want to change one near in my near blockchain and send it and swap it from a USDC in base blockchain to my MetaMask wallet. Okay. So, first of all, I will put one here. Then I will go to my MetaMask wallet. Ah, give me a second. H also here you can see all the tokens uh that we have available. H in this case I want to swap it for USDC. Here you can see

227. 1:45:00

     that we have a different chains ethereum, arbitrum, base, nosis, Solana etc. H I want to change it for USDC in base. Here it says the amount I will receive. And here in my MetaMask wallet h this is a test account that I just created and I have nothing h here. I will copy my address. I will paste it here and I will swap it. Okay.

228. 1:45:31

     H first the DAP is going to ask me to depositing my tokens. I will depose them. I will approve it. The transaction is execution executing right now. Then I will confirm the swap. I will sign a message and just in a couple of seconds the H intent will be approved. Now the intent is seted. Now I can go to my MetaMask wallet. Just in maybe 15 20 seconds my USDC are going to appear here.

229. 1:46:07

     Or if not then we go to a a base lo base base scan. Now you can see that my USDC appear. I have two and a half USDC. And here you can see to in the basis scan I have a 2.39 USDC. And that's all. Thank you, Peter, and thank you, Nirwick.

230. 1:46:43

     Thank you so much for coming, Thomas. And um just quickly want to say everyone, check out the what did you ship this week stream. It's going live in one hour. It's where Cameron from Near AI will take you through everything that's being shipped on AI wise in the new ecosystem. It's a weekly show and I want to give a huge shout out to all the teams that are building and to Cameron for highlighting all the good AI stuff.

231. 1:47:06

     It's 8:00 PM C. Make sure to tune in on the near protocol official X account. And um lastly, want to say apologize for this call taking so long today. We had so many things on the agenda. Um nearest where users own AI AI agents coordinate and the agentic internet scales. This community is grinding, building, and showing the world what's possible. So shout out to every developer, founder, and community contributor pushing the vision forward. Near isn't just a protocol. It's a movement for truly open

232. 1:47:36

     user owned AI. It's the future that we decides to build ourselves and we are all agreeing on to fork that. As always, what what has been presented on this uh town hall is not financial advice. Do your own research. Use your own judgment. But if you're paying attention, you know where the signal is. Subscribe to Newe to stay in the loop.

233. 1:47:56

     And if you're building something amazing, this stage is yours. Do not hesitate to hit us up. See you next month. Keep building, stay AI, and remember, the future of AI runs on Near. See you. Thanks for watching. Truly open AI is near.

234. 1:48:15

     Follow near protocol. Subscribe to Nearweek.