doctorambientJun 11
Adrian Sampson
seems insensitive of CACM to publicly post someone’s AI psychosis? https://cacm.acm.org/blogcacm/i-spent-a-year-asking-ai-to-solve-maths-hardest-problem-heres-what-happened/
May 31 at 7:32pmWeb
Show threadG. WozniakMay 31
@adrian Jebuz...
Show threadatsuzakiMay 31
@adrian fucking EUGH
Show threadsreeMay 31
@adrian Architecture and Hardware section? That said, I think this is no different than how cranks in physics.
Show threadeestileib (she/hers)Jun 1

@sree @adrian

I was going to say, this reminds me of being buttonholed at an alumni event by someone you've never met who wants to tell you about his perpetual motion machine.

Show thread/sbin/samMay 31
@adrian Oof.
Show threadSlava PestovMay 31

@adrian “Dmitry Khanukov is a seasoned entrepreneur and technology leader with over two decades of experience in building and scaling B2B tech products, AI platforms, and B2C marketplaces.”

This is just like when the ex-Uber CEO guy said he was on the verge of a breakthrough in quantum physics

Show threadAdrian SampsonMay 31
@slava why is the world so full of this one type of dude
Show threadG. WozniakJun 2
@adrian @slava Sorry to say, but it's because we are so desperate for measures of "success". And currently that metric goes to one's ability to make money.
Show threadMark DominusMay 31
@adrian Cruel, even.
Show threadsynlogic4242May 31
@adrian meanwhile, here in the real world, AI cant count the number of occurences of letter in a word
Show threadRob RicciJun 1
@adrian I have words to say about this but I need to turn them from assumptions into proven statements
Show threaddrmorrJun 1
@ricci @adrian "I have discovered a marvelous proof of this theorem but I ran out of tokens"
Show threadHenrý ÓlsonJun 1
@adrian think the blog post itself might be the work of an llm too \:
Show threadNicola Elle 🏳️‍⚧️Jun 1

@adrian

🤦‍♀️ 🤦‍♀️ 🤦‍♀️

I facepalmed so hard that I accidentally generated a wormhole into my own forehead.

Show threadSam Tobin-HochstadtJun 1
@adrian I have bad news:
Show threadbhaktiJun 1
@samth @adrian lol I guess at least there’s a record now
Show threadDaniel GibsonJun 1
@samth @adrian
good to know that journal editors still don't take blogs seriously
Show threadYannick ForsterJun 1
@samth @adrian thank you for flagging this issue, I would not have been able to stay as calm after this outrageous response
Show threadSam Tobin-HochstadtJun 1
@yforster @adrian I deleted several previous drafts of that tweet.
Show threadLindsey KuperJun 2
@samth @yforster @adrian seriously, thank you for speaking up, Sam, and for finding the right words to do it
Show threadJon SterlingJun 2
@samth @adrian The really funny thing is, I was saying to Carlo, "Wow this is so bad, a senior person needs to step in and respond to this, perhaps Moshe Vardi?" and then he sent me your thread 🤦‍♂️ 😵‍💫
Show threadSam Tobin-HochstadtJun 2
@jonmsterling @adrian yes we were on the way to lunch when he delivered the bad news to you
Show threadJon SterlingJun 2
@samth @adrian lmao
Show threadSam Tobin-HochstadtJun 2
@jonmsterling @adrian in fact the whole thing is his fault; he saw Adrian's post, DMed it to me, and then I got mad online at Moshe.
Show threadTristan NguyenJun 2
@jonmsterling Vardi’s response is arrogant, disrespectful, and distasteful
Show threadJon SterlingJun 2
@TristanNguyen I’m really shocked by it…
Show threadTristan NguyenJun 2
@jonmsterling I was shocked when I first read the article, which was seemingly composed with the help of a chatbot (the most technical part).
Show threadG. AllaisJun 2
@samth @adrian this kind of attitude cannot possibly be compatible with the ACM code of ethics
Show threadG. WozniakJun 2
@samth @adrian That's some serious "Dig up, stupid!" vibe.
Show threadThe Audacity of NopeJun 3
@samth @adrian well shit
Show threadTobiasJun 7
@samth @adrian 😱
Show threadAlexandooJun 1
@adrian draw the rest of the owl
Show threadDaniel McLauryJun 3

@adrian Seems like his actual claim is that he has created computer-readable versions of published articles on P != NP, including proofs of the form "If X is true, then P != NP."

Does not seem like he claims this has lead to any new insight or any new results, just that he got the lean files to compile. Which, sure, I can believe that someone did that.

When he says "I have a complete formalization of the P ≠ NP problem" -- emphasis on "problem" -- I guess that literally means that he has a computer-readable statement of the *proposition* "P != NP"

Only thing really suspect here is the one highlighted sentence where he apparently thinks that this is a first step towards making actual progress on the problem.