My new project: thinking through and refining the concept of "an AI-shaped problem". AI has shocked us, both by doing things we had considered beyond its abilities and making baffling blunders in things that should have been easy. I believe this shock to be a mismatch ...
1
21
.. between expectations and observation, and a remedy is to have better expectations. The first post in the series: amahabal.substack.com/p/ai-s…
14
Abhijit Mahabal retweeted
Newly elected Senator Abdul El-Sayed has an H-Index of 37. This makes me inordinately happy. Expertise matter. Science matters. Facts matter. Rationality is a magic power -- @AbdulElSayed
#ScienceMatters #FactsMatter #BLUEWAVE
Abhijit Mahabal retweeted
Replying to @BachFrancis
(1/7) I'd like to provide some additional context for the remarks that can be found in this interview.
(2/7) My discomfort with the attention being paid to existential risk---and with the ongoing dominance of the public discussion of such risks by a few so-called thought leaders---is that there's an elephant in the room that is being missed. That elephant is the effect of all of this discussion on young people, and indeed the effect---beginning with social media a decade ago---of information technology on young lives broadly speaking.
(3/7) I recently read an interview by Jill Lepore in which she reports on discussions that she had recently with a group of college students: "They had just three things to say. One, social media completely destroyed their childhoods. Two, AI is destroying their college education and their likely prospects for employment. And three, what the hell are we all doing?" To add a fourth bullet that I've heard in similar discussions: "I'm not going to have children, because the existential risk is just too high."
(4/7) Please contemplate such a statement, those of you who think that it's OK that the public, widely diffused, discussion coming from seemingly authoritative figures is so adamant about existential risk.
(5/7) Let me add that I have little interest in a debate about what "superintelligence" might mean, or whether we're currently seeing the emergence of superintelligence. For what it's worth, I would certainly agree that the solution of extremely challenging math problems (for example) exhibits a very impressive form of intelligence, although I would want to argue that such intelligence is impossible to dissociate from the collective of humans that provided data to the machine learning system (which itself has become a kind of collective).
(6/7) Such philosophical debates are entertaining and even important, but when they then head in the direction of "when such superintelligence emerges, there's a good chance that it will want to kill all humans", I say "science fiction". Again, I don't think that it's bad to consider such thoughts, but I strongly believe that it's bad for society, and especially young people, for such thoughts, unaccompanied by any serious real-world analysis, to become a dominant form of discourse of the era.
(7/7) We are damaging human lives, and the consequences of that damage will no doubt be a major theme of the next several decades.
Abhijit Mahabal retweeted
AI has the potential to wipe out the human race, but there are also downsides.
Abhijit Mahabal retweeted
Replying to @ns123abc
AI will cure cancer if you input the cure into the prompt window
Abhijit Mahabal retweeted
What is mind-boggling here is that googling for one of the possibly fake citations that R2 highlighted gives an AI overview that makes it look like the paper existed, but with a circular reference back to the @iclr_conf paper citing it.. A glimpse of our afactual world to come..
Given today’s AI inundation, I just realized, an old pun (two meanings) has now become a double pun: On being asked “can you pun on the zodiac?” the punster replied “By Gemini, I can sir”.
50
Wonderful exploration of a math problem using several LLMs and why it ain’t slam dunk that they will accelerate research
Let me sum up the episode that took place in this thread, because I think it's instructive and microcosmic of at least one way that I expect LLM-"assisted" research to progress in the real world. Anecdotal and you know I'm predisposed as a hater but I think it's a good case study
76
As our reliance on ChatGPT grows, we need to know when it will be helpful and when it might hallucinate. Large language models are still new, and our understanding of their blind spots is limited, but philosophers and cognitive scientists have long studied human knowledge. 🧵
1
75
Drawing on a framework proposed by British philosopher A. C. Grayling, this post analyzes common LLM failure modes and offers practical tips for predicting whether an LLM will help or hinder in a given situation. amahabal.substack.com/p/anti…
51
I have a huge trade deficit with Amazon. It has been ripping me off. From now on, I am going to require Amazon to pay me money if they want to continue shipping stuff to me. No deal until I reach zero deficit!!!
I wonder if, on the average, folks who pay OpenAI $200 a month are more gullible. Perhaps a tad less likely to look carefully enough at the output? Some cognitive dissonance because of the sunk capital?
⚠️⚠️⚠️ model collapse warning & a caution for science ⚠️⚠️⚠️
Scary how AI influencers are claiming “hallucinations are over” and Deep Research is revolutionizing science when actual reports by people who looked carefully post reports like the one below.
And @hoofball_69 100% nails the serious problem that fans like @DeryaTR_ tend to overlook: a lot of these errors are difficult to detect and will seep into the literature. GenAI will not be able to undo the mess.
Model collapse for science is no laughing matter. But it is now imminent.
1
173
Why is today's AI such a baffling mix of genius and stupidity? I explore one key aspect: open.substack.com/pub/amahab…
1
58
















