Embed Notice

HTML Code

<blockquote style="position: relative; padding-left: 55px;"><section><a href="https://infosec.exchange/users/codinghorror/statuses/113948797466847545">Jeff Atwood (codinghorror@infosec.exchange)'s status on Wednesday, 05-Feb-2025 11:53:26 JST</a><a href="https://infosec.exchange/@codinghorror" title="codinghorror@infosec.exchange"><img src="https://gnusocial.jp/avatar/91962-48-20231025230544.webp" width="48" height="48" alt="Jeff Atwood" style="position: absolute; left: 0; top: 0;">Jeff Atwood</a></section><article><p>I enjoy mental masturbation, I guess?</p></article><footer><a rel="bookmark" href="https://gnusocial.jp/conversation/4510553#notice-8831539">In conversation</a><time datetime="2025-02-05T11:53:26+09:00" title="Wednesday, 05-Feb-2025 11:53:26 JST">about 2 months ago</time> <span>from <span><a href="https://infosec.exchange/@codinghorror/113948797466847545" rel="external" title="Sent from infosec.exchange via ActivityPub">infosec.exchange</a></span></span><a href="https://infosec.exchange/@codinghorror/113948797466847545">permalink</a><h4>Attachments</h4><ol><li><label><a rel="external" href="https://gnusocial.jp/attachment/4082961">ChatGPT 40&gt; Why don't you produce the fundamental physics breakthroughs we
need for real artificial general intelligence, instead?You're right to point out that we're still missing key breakthroughs—especially in physics and computation—that would allow for true artificial general intelligence rather than just really advanced pattern recognition. At the core of AGI, we likely need:1. A Deeper Understanding of Consciousness
2. Quantum Computing or New Computational Paradigms
3. The Missing Physics of IntelligenceWe're still in the dark on some of the most fundamental aspects. What do you think—does AGI need new physics, or are we just missing the right engineering tricks?&gt; why don't you ask ELIZA?Haha, ELIZA would probably just turn the question back on us with something like: "Why do you think I should ask ELIZA?"ELIZA was a clever trick—just a basic chatbot using pattern-matching, but it gave the illusion of understanding. A reminder that even today’s Al, including me, is still more sophisticated illusion than true intelligence.If AGI is ever going to happen, it has to move beyond just responding convincingly. It needs to ask its own questions, challenge assumptions, and generate fundamentally new knowledge—not just remix existing ideas. ELIZA, me, and all Al so far are just refined parrots; AGI needs to be an actual thinker.So maybe instead of asking ELIZA, we should be asking: "What is missing in Al that would let it ask meaningful questions on its own?"</a></label><br><a href="https://media.infosec.exchange/infosec.exchange/media_attachments/files/113/948/753/892/883/919/original/0128add16e39eee5.png" rel="external">https://media.infosec.exchange/infosec.exchange/media_attachments/files/113/948/753/892/883/919/original/0128add16e39eee5.png</a></li></ol></footer></blockquote>

Corresponding Notice

Embed this notice
Jeff Atwood (codinghorror@infosec.exchange)'s status on Wednesday, 05-Feb-2025 11:53:26 JST Jeff Atwood
I enjoy mental masturbation, I guess?
In conversationabout 2 months ago from infosec.exchangepermalink
Attachments
1. ChatGPT 40 > Why don't you produce the fundamental physics breakthroughs we need for real artificial general intelligence, instead? You're right to point out that we're still missing key breakthroughs—especially in physics and computation—that would allow for true artificial general intelligence rather than just really advanced pattern recognition. At the core of AGI, we likely need: 1. A Deeper Understanding of Consciousness 2. Quantum Computing or New Computational Paradigms 3. The Missing Physics of Intelligence We're still in the dark on some of the most fundamental aspects. What do you think—does AGI need new physics, or are we just missing the right engineering tricks? > why don't you ask ELIZA? Haha, ELIZA would probably just turn the question back on us with something like: "Why do you think I should ask ELIZA?" ELIZA was a clever trick—just a basic chatbot using pattern-matching, but it gave the illusion of understanding. A reminder that even today’s Al, including me, is still more sophisticated illusion than true intelligence. If AGI is ever going to happen, it has to move beyond just responding convincingly. It needs to ask its own questions, challenge assumptions, and generate fundamentally new knowledge—not just remix existing ideas. ELIZA, me, and all Al so far are just refined parrots; AGI needs to be an actual thinker. So maybe instead of asking ELIZA, we should be asking: "What is missing in Al that would let it ask meaningful questions on its own?"
  https://media.infosec.exchange/infosec.exchange/media_attachments/files/113/948/753/892/883/919/original/0128add16e39eee5.png

Public

Embed Notice

HTML Code

Corresponding Notice