OpenTube

The Real Story Behind OpenAI’s New Astra Model

Cal Newport

42,531 views

Cal Newport takes a critical look at recent AI News.

More from Cal

Download Cal’s FREE guide to cultivating a deeper life: calnewport.com/ideas
Learn more about Cal’s books: calnewport.com/books
Listen to Cal’s podcast: thedeeplife.com/listen

Chapters

(0:00) Does Open AI’s Astra mean AGI has arrived?
(3:07) What actually happened?
(11:35) What does this mean for mathematics?
(21:35) What does this mean for OpenAI?

Resources Mentioned:

Buy Cal’s latest book, “Slow Productivity” at www.calnewport.com/slow
https://x.com/polynoamial/status/2083...
https://x.com/deanwball/status/208354...
https://x.com/kevinroose/status/20836...
https://x.com/mattshumer_/status/2083...
https://x.com/polynoamial/status/2083...
https://x.com/__alpoge__/status/20838...
https://x.com/polynoamial/status/2083...
https://x.com/thomasfbloom/status/208...

Sponsors:

https://www.donedaily.com

Credits:

Podcast Production: Jesse Miller
Newsletter/Research: Nate Mechler
Theme Music: Jay Kerstens

20 Comments

@9ZweihandeR9 2 months ago

Wow, this program did a good job at the specific task it was optimized for. I bet it's going to be great at every other task imaginable!

@SavageCommentaryOriginal 2 months ago

I use frontier models everyday for complex tasks. How anyone can believe these things are anywhere near approximating sentience is just mindboggling. If you have even a vague understanding of how they work it's laughable.

@lemuelbach 2 months ago

I appreciate you still making these videos, Cal!!

@UserProfile-e9b 2 months ago

The biggest issue I have is that sentience really won’t matter. As long as young people start using LLMs to replace their own thinking they will slowly take over our ability to think for themselves. Long term we’re messed up and will fully rely on the LLM, which will destroy creativity and fast track us into a dystopian world where humans think everything that comes from LLMs is the truth with no ability to fact check it.

@akkanomark 2 months ago

That shit is pure desperation to keep the cassino rolling

@RAFMnBgaming 2 months ago

1:43 I think the most unbelievable bit here is the idea that the cost of AI will fall.

@krox477 2 months ago

Thank you prof for breaking down complex matters in simple words

@amarug 2 months ago

My split personality is really pushed to the edges. I am sooo excited about these new tools and watching the progress and yet soooooo tired of the hype train where every new model "just won the whole AI race and ushers in a new world order" 😭

@cherubin7th 2 months ago

The software like mathematica did more "discoveries" because now you have software that can just search the symbol space for a solution.

@ahm6r 1 month ago

Mr Cal , I Have just finished Your book (Deep Work) , I just want to say THANK YOU

@AquilaLabsCA 2 months ago

According to Altman we're now in the singularity, which sounds awesome to me, because I went totally blind years ago due to a mitrochrondial genetic defect. Considering we're in the singularity now, I'm sure it's just another week or two out before we know how to regenerate optic nerves, right? Cool, I can't wait to see again.

@queasybeetle 2 months ago

LLM is the ultimate Dunning Kruger machine.

@TheLizardKing752 2 months ago

I'm glad Deep Questions is back

@sumemr2081 2 months ago

Been eagerly awaiting this video

@klnmn3722 2 months ago

As soon as I saw this story, I was wondering what you’d say about it! Excited to hear your thoughts

@CliffNickerson 2 months ago

17:15 This is a fantastic metaphor! Bravo, sir!

@clickgear9407 2 months ago

His "AI is a force that gives us meaning crowd" line is devastating and so true. If you listen to Peter Diamandis and others in his immediate orbit, it seems clear that AI is their literal religion. So many seem to have invested their metaphysical hopes in AI to make them immortal. Sorry, God can't be supplanted, and he certainly can't be manufactured by people.

@swyveu 2 months ago

overextrapolation really is one of the most detrimental cognitive dysfunctions of the human mind,
especially when you only apply it to things that are favorable to you.

@moralmasochist1 2 months ago

The goat. Do a video on Joseph Gordon Levitt; he called you out!

@HA-dw5ik 2 months ago

I think pouring cold water on every new development misses the most important thing. Models are steadily improving as expected, as they have been for years, certainly not at the same pace for every field, but steadily in every field nonetheless.