J1 Races Are Here, Is AGI actually possible? And the DoD's plan for AGI

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

1. 0:00

   Good morning Legion. J1 is back. It's happening. And it's happening sooner than most of us expected. Race registration opened yesterday and closes today in about five hours time, with the first race slated to start on tomorrow, Friday afternoon eastern time.

2. 0:16

   This is the first race of J1 Championship season. There will be many more to come, so if you feel a little rushed right now, don't worry, you have plenty of time to make most of the races as we kick off. Our first story today is a bold thesis from Tim Dettmer, a research scientist at Allen AI and incoming professor at Carnegie Mellon.

3. 0:34

   Tim's blog is a harsh critique on the supposed sloppy thinking in the Bay area as it relates to AGI. As he writes in summary. AGI is commonly conceived will not happen because it ignores the physical constraints of computation.

4. 0:48

   The exponential costs of linear progress, and the fundamental limits we are encountering. Superintelligence is a fantasy because it assumes that intelligence can recursively self improve without bound, ignoring the physical and economic realities that constrain all systems.

5. 1:03

   These ideas persist not because they're well founded, but because they serve as compelling narratives in an echo chamber that rewards belief over rigor. Be sure to read his entire deep dive to see what his implicit assumptions are, and how these might change over time.

6. 1:18

   Our second story today is in the opposite direction. In a groundbreaking development, lawmakers have introduced a provision in the fiscal 2026 National Defense Authorization Act that mandates the establishment of an artificial intelligence futures steering committee within the US Department of Defense.

7. 1:34

   This committee will include future planning for artificial general intelligence focused on mitigation strategies, risks and opportunities in this domain. In other words, even if technologists are skeptical about AGI, the US Department of Defense isn't taking any chances.

8. 1:50

   Our third and final story is about a big time investment from the Astera Institute, and they're trying to solve one of the major primitives for AGI, Astera Neuro. A new neuroscience research effort aims to unravel how the brain transforms sensory inputs into conscious experience, with the hope of being able to use these discoveries for building AGI.

9. 2:11

   The initiative will be funded with $600 million over the next ten years. Overall, there's a lot happening in the run up to the end of 2025, in the NEAR ecosystem. Don't miss out on Legion announcements this week. We do have a couple of goose NFTs to give away for J1. And tune back in with us right here. We'll do this all again real soon.