▲ 380 ▼ Quote of the day by computer scientist Judea Pearl: 'Deep learning has instead given us machines with truly impressive abilities but no intelligence' — capturing the central fault line in modern AI. (www.techradar.com) submitted 21 hours ago by eicker@lemmy.world to c/technology@lemmy.world 64 comments fedilink hide all child comments
[–] emmy5482 3 points 6 hours ago (4 children) What are the chances? I don't know, but neither do you. However, it is non-zero Because they've already done it. https://www.forbes.com/sites/sandycarter/2026/09/10/did-openai-steal-a-1m-math-proof-four-questions-for-every-ceo/ permalink fedilink source parent hideshow 4 child comments replies: [–] robert@piefed.ca 2 points 3 hours ago* (3 children) This is an interesting story but like I said the sheer volume of these proofs is astonishing. If it was just that one particular proof then you’re above link would definitely be reason for skepticism but we’re talking about nearly 300 proofs here, all dumped at once. Many of them have been open problems for decades if not a century plus, and many of them are so significant that even solving one would have been math news of the decade. A scenario where all these breakthroughs just so happen to be made simultaneously, just in time for OpenAI to steal them is just not believable. That many significant breakthroughs coincidentally happening all at once is unprecedented in intellectual history. I understand that people hate AI, but the fact that people are denying that these breakthroughs in AI actually happened just shows that the Lemmy/Piefed users have, in general, a very minimal understanding of the relevant mathematical literature. You guys can downvote me all you want for saying that but that’s the truth. permalink fedilink source parent hideshow 3 child comments replies: [–] emmy5482 -1 points 2 hours ago (2 children) Many of them have been open problems for decades if not a century plus, and many of them are so significant that even solving one would have been math news of the decade. That isn't quite true. Many have been open for a long time. But most have paltry prizes and people just don't have the time or inclination to solve many of them Earlier releases of the solutions also were attributed to people who wrote papers, just not in English. Even then, solving them isn't really that big a deal. This is an attempt at a PR stunt to try and fix one of the biggest problems with ai. That they can't do math. You can see extremely famous videos of people asking LLMs how many e's in seventeen and being told it's actually 2 words. permalink fedilink source parent hideshow 2 child comments replies: [–] robert@piefed.ca 2 points 2 hours ago (1 child) This comment is riddled with verifiable, factual inaccuracies and does nothing but demonstrates your own ignorance and misunderstanding. I don’t even know how to engage with someone such as yourself who is not grounded in reality. It is like trying to engage with a climate change denier permalink fedilink source parent hideshow 1 child comment replies: [–] Feathercrown@lemmy.world 2 points 1 hour ago Yeah I thought they had some reasonable opposition but this comment proves they don't really know what they're talking about unfortunately permalink fedilink source parent
[–] robert@piefed.ca 2 points 3 hours ago* (3 children) This is an interesting story but like I said the sheer volume of these proofs is astonishing. If it was just that one particular proof then you’re above link would definitely be reason for skepticism but we’re talking about nearly 300 proofs here, all dumped at once. Many of them have been open problems for decades if not a century plus, and many of them are so significant that even solving one would have been math news of the decade. A scenario where all these breakthroughs just so happen to be made simultaneously, just in time for OpenAI to steal them is just not believable. That many significant breakthroughs coincidentally happening all at once is unprecedented in intellectual history. I understand that people hate AI, but the fact that people are denying that these breakthroughs in AI actually happened just shows that the Lemmy/Piefed users have, in general, a very minimal understanding of the relevant mathematical literature. You guys can downvote me all you want for saying that but that’s the truth. permalink fedilink source parent hideshow 3 child comments replies: [–] emmy5482 -1 points 2 hours ago (2 children) Many of them have been open problems for decades if not a century plus, and many of them are so significant that even solving one would have been math news of the decade. That isn't quite true. Many have been open for a long time. But most have paltry prizes and people just don't have the time or inclination to solve many of them Earlier releases of the solutions also were attributed to people who wrote papers, just not in English. Even then, solving them isn't really that big a deal. This is an attempt at a PR stunt to try and fix one of the biggest problems with ai. That they can't do math. You can see extremely famous videos of people asking LLMs how many e's in seventeen and being told it's actually 2 words. permalink fedilink source parent hideshow 2 child comments replies: [–] robert@piefed.ca 2 points 2 hours ago (1 child) This comment is riddled with verifiable, factual inaccuracies and does nothing but demonstrates your own ignorance and misunderstanding. I don’t even know how to engage with someone such as yourself who is not grounded in reality. It is like trying to engage with a climate change denier permalink fedilink source parent hideshow 1 child comment replies: [–] Feathercrown@lemmy.world 2 points 1 hour ago Yeah I thought they had some reasonable opposition but this comment proves they don't really know what they're talking about unfortunately permalink fedilink source parent
[–] emmy5482 -1 points 2 hours ago (2 children) Many of them have been open problems for decades if not a century plus, and many of them are so significant that even solving one would have been math news of the decade. That isn't quite true. Many have been open for a long time. But most have paltry prizes and people just don't have the time or inclination to solve many of them Earlier releases of the solutions also were attributed to people who wrote papers, just not in English. Even then, solving them isn't really that big a deal. This is an attempt at a PR stunt to try and fix one of the biggest problems with ai. That they can't do math. You can see extremely famous videos of people asking LLMs how many e's in seventeen and being told it's actually 2 words. permalink fedilink source parent hideshow 2 child comments replies: [–] robert@piefed.ca 2 points 2 hours ago (1 child) This comment is riddled with verifiable, factual inaccuracies and does nothing but demonstrates your own ignorance and misunderstanding. I don’t even know how to engage with someone such as yourself who is not grounded in reality. It is like trying to engage with a climate change denier permalink fedilink source parent hideshow 1 child comment replies: [–] Feathercrown@lemmy.world 2 points 1 hour ago Yeah I thought they had some reasonable opposition but this comment proves they don't really know what they're talking about unfortunately permalink fedilink source parent
[–] robert@piefed.ca 2 points 2 hours ago (1 child) This comment is riddled with verifiable, factual inaccuracies and does nothing but demonstrates your own ignorance and misunderstanding. I don’t even know how to engage with someone such as yourself who is not grounded in reality. It is like trying to engage with a climate change denier permalink fedilink source parent hideshow 1 child comment replies: [–] Feathercrown@lemmy.world 2 points 1 hour ago Yeah I thought they had some reasonable opposition but this comment proves they don't really know what they're talking about unfortunately permalink fedilink source parent
[–] Feathercrown@lemmy.world 2 points 1 hour ago Yeah I thought they had some reasonable opposition but this comment proves they don't really know what they're talking about unfortunately permalink fedilink source parent