I think we are still a long way off with robotics, it's like every generation keeps promising Rosie from the Jetsons but the best we can do is a vacuum bot that gets stuck and a litter box that can clean itself.
I think we are still a long way off with robotics, it's like every generation keeps promising Rosie from the Jetsons but the best we can do is a vacuum bot that gets stuck and a litter box that can clean itself.
Mark Zuckerberg is not the hero we wanted but the hero we needed. Good on him for making Llama 3 open source.
Randy Travis sings again, courtesy of AI.
The past is a foreign country; they do things differently there.
Jesus Christ. GPT4.o is your new best friend.
https://twitter.com/hey_madni/status...99252607324629
And Google responds. Mind blown
https://twitter.com/i/status/1790435235433238829
The exponential growth of AI's capabilities is hard to grok, even when they tell you about it explicitly. Shit is going to get crazy much sooner than we are ready for it to.
https://www.nytimes.com/2023/07/21/p...fake-love.html
ride bikes, climb, ski, travel, cook, work to fund former, repeat.
Ilya leaving OpenAI is very very not good.
What could go wrong?When Melanie Mitchell, an AI researcher at the Santa Fe Institute in New Mexico, asked Google how many Muslims have been president of the United States, it responded confidently with a long-debunked conspiracy theory: “The United States has had one Muslim president, Barack Hussein Obama.”
Seems like there's a supervillain origin story there: smart nerds, deeply confused by the mystery of popularity, set out to build their own friends. Attach such foundational importance to popularity that their new friends are never able to define truth beyond "number of mentions."
<p dir="rtl">
Make efficiency rational again</p>
AI could start getting things right, not wrong. 100 percent accuracy is something no human can achieve. If an AI were to achieve accuracy instead of narratives capable of slipping past human reason then that would be godlike. However impressive, AI is not yet close to achieving such a high standard. AI is maybe 80-percent accurate when it comes to basic search & inquiry. The last 20, 10, 5, percent is orders of magnitude harder to achieve than today's probabilistic 80 percent. Actual reasoning, an AI model that can verify it's own results, will require something other than today's LLMs.
Forum Cross Pollinator, gratuitously strident
And yet half the people can’t even tell it’s not human
https://arxiv.org/abs/2405.08007
Forum Cross Pollinator, gratuitously strident
that poor horse...
Forum Cross Pollinator, gratuitously strident
lol, a "Miss AI" beauty pageant. $20k in prizes will go to creators of AI-generated models. "Contestants" must not only submit photos, but answer the pageant questions like "how would you make the world a better place?"
https://www.euronews.com/culture/202...enerated-women
Zis fellow Görman iz a zmart won!
https://youtu.be/pZVngnr8KN4?si=wwPJIaQzAfNmHSiL
It's a war of the mind and we're armed to the teeth.
The past is a foreign country; they do things differently there.
wait .......
Wut?!??!?
Bookmarks