Comments by "" (@DaveShap) on "Wes Roth" channel.

  1. Elon Musk: "Hold my beer"
    468
  2. It's getting spicy!
    448
  3. Finally! I've been advocating for autonomy for a while. This was always inevitably the way forward. But it's going to be economically disruptive, and potentially painful. I'm still working on understanding how this will all play out...
    281
  4. Good job getting to the "one video a day" scheme. Don't burn out!
    250
  5. What still gets me is that AI agents just became a thing but we are NO WHERE near the ceiling of model capability and people are already planning "oh this is as far as it's going to go" - no it ain't!
    227
  6. I was literally just thinking last night as I was going to bed "If a bunch of us independent people are building AI agents, then you know darn well the tech giants are." I feel like 2024 is the year that jobs start evaporating fast.
    182
  7. When people tell me that they think their job as a plumber is safe, I'm not so sure.
    120
  8. If all this is true then this is what Sam meant when he said "the era of gigantic models is over"
    108
  9. It's happening!!!!
    76
  10. Thanks for the shout-out!
    51
  11. Dang dude, churning it out. Got your finger on the pulse! Doing humanity's work, brother. Thank you, and keep it up!
    44
  12. Keep the conversation going!
    36
  13. Millions of disembodied robot hands endlessly flipping cubes is a very odd sort of ring of Dante's Inferno. What did they do wrong? Generate political misinformation?
    28
  14. Grok2 better than GPT4 in May... but also AI surpassing all humans within 5 years... interesting
    26
  15. "A few minutes to kill" - well done
    18
  16. That Ilya, Sam, Elon, Greg, and whoever else made the unilateral decision about what's best for everyone else, without consulting... anyone? This is increasingly catastrophic. Considering Ilya's egregious lack of judgment with the Sam ousting, this is not someone who has demonstrated the wherewithal to make strategic decisions. Scientists and engineers have their role. Global safety and strategy is not one they should be playing.
    12
  17. Doing important work. Keep it up!
    12
  18. It will be autonomous! Glad to see NVIDIA taking the lead on this.
    9
  19. I've been telling people for the last year DO NOT COMPETE WITH OPENAI. It's honestly why I have declined getting involved with any startups, they're all gonna get nuked.
    7
  20. Sovereign AI as defined by Jensen is congruent with "structural realism" e.g. all nations will seek power as a deterrent. Democratizing access to AI across nations will help maintain a power status quo.
    7
  21. Very typical trajectory for a cult of personality. It's more about Sam's ego than anything else now.
    6
  22. Great work!
    4
  23. shockedpikachu.jpg
    3
  24. History will record 2024 as the year it all really began.
    3
  25. Great research thanks.
    3
  26. Btw, I call the theory here "Above the canopy" e.g. we get out of the dark forest by going up, not down. Like when Bilbo climbed up the tree to get his bearings out of Mirkwood. (That's where I got the idea). Get away from the noxious fumes and the spiders by getting above the noise and nonsense.
    3
  27. Yann's position is often unclear but in this case I completely agree with him.
    3
  28. Thanks for the deep dive
    2
  29. Neurobiological failures of humans are more to blame. The algorithm just surfaces our weaknesses.
    2
  30. The fact that the investigation ended right after Elon filed a lawsuit, and that OpenAI responded with a blog post means they are in more dire straits than they are letting on.
    2
  31. This new AI-generated aesthetic is called "new surrealism"
    2
  32.  @Rolandfart  never underestimate an autistic systems thinker!
    2
  33. Oh hey thanks for the shoutout. Cheers!
    2
  34. 9:25 - we have the concept of a plan...
    2
  35. What do you use for screen capture?
    2
  36. Comparing an LLM to bag-of-words ML is the strangest cope I've seen yet. Re: Joscha Bach
    2
  37. Ready for my daily dose of voltage.
    1
  38. DARPA PLZ NO DISAPPEAR ME ~ Wes Roth, 2024
    1
  39. Great research
    1
  40. FYI the big Q* leak was a total fabrication.
    1
  41. Glad you're face is back on the screen. Adds a lot to the experience.
    1
  42. Guys, look up Holographic Universe Theory and then tell me how much your noodle is cooked.
    1
  43. I don't sleep when there's a big project to work on... ☠
    1
  44. "agenticness" is a strange way of pronouncing "autonomous"
    1
  45. I dunno man. I'm a paid subscriber to ChatGPT and have even been using the API and ChatGPT seems to be getting dumber and more useless. Claude is now easily 10x less effortful to use.
    1
  46. OODA loop = TRC (Terminal Race Condition). Thanks for the good video.
    1