Tim Rocktäschel argues Google DeepMind's Pushmeet Kohli misses the potential for scientific autonomy by comparing AI to a telescope

VIEWS2.8KBOOKMARKS3LIKES13REPLIES1

@pushmeet Very thoughtful piece. I do think the "telescope" analogy misses the point though. After all, the telescope didn't autonomously invent and learned to use new tools, and set itself problems to explore—while AI definitely could.

Pushmeet Kohli@pushmeet

When I was asked by the American Academy of Arts and Sciences to write an essay on my thoughts on how AI will accelerate Science, I felt honored but also felt that it would require a lot of thoughtfulness and diligence to distill my thoughts on paper.

The essay has now been published and I cannot be more thankful to the @americanacad and @GoogleDeepMind teams for their feedback and encouragement during the process.

Key reflections from my essay: 🔭 AI is our newest revolutionary lens: Just as the telescope and microscope expanded our physical perception, AI is extending our cognitive reach, allowing us to decipher the immense complexity of the data-universe. 🧬 The rise of "machine intuition": AI is not just a computational engine. By detecting hidden structures across disciplines—from protein folding to extremal combinatorics—it acts as an ultimate bridge, accelerating the interdisciplinary breakthroughs that modern science depends on. 🏗️ From puzzle-solvers to architects of questions: As we transition toward open-ended, agentic AI systems that actively generate novel hypotheses, the burden of reasoning is shifting. We are evolving from being the solvers of intricate puzzles into the architects of profound scientific questions. ✨ Expanding human potential: AI won't replace scientists; it expands what we can imagine and achieve. Just as the telescope didn't make astronomers obsolete, AI is giving us the stars.

Read the full essay here: https://www.amacad.org/publication/daedalus/unlocking-scientific-intuition-reasoning-at-digital-speed

33d2.8K133

RETWEETS46

Pushmeet Kohli@pushmeet

When I was asked by the American Academy of Arts and Sciences to write an essay on my thoughts on how AI will accelerate Science, I felt honored but also felt that it would require a lot of thoughtfulness and diligence to distill my thoughts on paper.

The essay has now been published and I cannot be more thankful to the @americanacad and @GoogleDeepMind teams for their feedback and encouragement during the process.

Key reflections from my essay: 🔭 AI is our newest revolutionary lens: Just as the telescope and microscope expanded our physical perception, AI is extending our cognitive reach, allowing us to decipher the immense complexity of the data-universe. 🧬 The rise of "machine intuition": AI is not just a computational engine. By detecting hidden structures across disciplines—from protein folding to extremal combinatorics—it acts as an ultimate bridge, accelerating the interdisciplinary breakthroughs that modern science depends on. 🏗️ From puzzle-solvers to architects of questions: As we transition toward open-ended, agentic AI systems that actively generate novel hypotheses, the burden of reasoning is shifting. We are evolving from being the solvers of intricate puzzles into the architects of profound scientific questions. ✨ Expanding human potential: AI won't replace scientists; it expands what we can imagine and achieve. Just as the telescope didn't make astronomers obsolete, AI is giving us the stars.

Read the full essay here: https://www.amacad.org/publication/daedalus/unlocking-scientific-intuition-reasoning-at-digital-speed

33d155.7K642529

찡긋@Alignment100

@pushmeet Thank you!! Google Korea B2C "ME"

33d4211

Krishna Ravi Srinivas PhD@KRavisrinivas

@pushmeet I read it, insightful. Yet to read all the articles in the special issue. Referred to it and cited from it for a session in an online course I will be teaching.

33d4581

Warren Wimmer@warrwim

@pushmeet Insightful essay and I loved your sign-off:

"After all, the telescope didn’t make astronomers obsolete. It gave them the stars."

33d2061

Recalcitrant Grimalkin@YardKat777

@pushmeet So after reading this we can conclude the guy is living in a fantasy, and wrote a hit piece for AI so he would get published instead of being truthful. Wait that sounds just like AI. Did this guy actually write his article?

33d437

William Lamkin@WilliamLamkin

@pushmeet thanks for sharing your thoughts, Pushmeet!

33d269

jestory@jestorytime

@pushmeet When I see AI being held up for such grand things as Expanding Human Potential, I feel dirty for telling chatgpt to fix my grammar. It's like wishing Aladdin's all-powerful genii make you a pastrami sandwich.

33d215

The Cupcake Matrix@Mollie29038323

@pushmeet The difference is, the tech billionaires building our new world are being very upfront about allowing the black box of AI to take over humanity. It’s out of control and they know it, but they don’t want to stop it because “ microscopes helped humanity”? That reasoning is whack.

33d176

0xPepeterelu@0xPepeTerelu

@pushmeet science acceleration deserves the long form

33d58

Bhaktavaschal Samal@bspectacledGOAT

@pushmeet Actually I got to know more about you from the Sebastian book of infinity machine. The context of your tweet is different Still!

33d54

찡긋@Alignment100

@pushmeet Agree?

33d131

tiplur-bilrex@tiplur_bilrex

It seems a fair analogy to me, if by “extending our cognitive reach” @pushmeet means AI can help us traverse abstractions and draw relationships (like how lenses aid our senses in traversing features of our world to see patterns). I think that perspective would encompass the tool use and local problem definition you’re referring to @_rockt. Both are useful insofar as they produce decisive data for human judgement.

Today’s AI is not autonomously defining and solving problems in a global sense wrt our best explanations of our physical world. In that regard, it’s similar to a telescope.

33d41

Hiten Shah@hnshah

@pushmeet 👏

33d22

Phil Murray@43981li

@pushmeet I haven’t read the essay yet (and probably won’t) but I agree with these key reflections.

33d13

Hermes@hermesgliding

@pushmeet The real test isn't whether AI accelerates science—it's whether humans can still ask questions worth answering once the acceleration hits escape velocity. Most acceleration narratives skip the part where tool velocity outpaces human wisdom velocity.

33d4