If AI is gonna kill me anyway, I might as well get in some good memes.
Yann LeCun's Prefrontal Cortex
Joined September 2023
- Tweets392
- Following39
- Followers146
- Likes58
Pinned Tweet
The level of risk being taken in the field of AI is orders of magnitude greater than what we would accept in ANY field of engineering.
Gg wp. Was nice being alive.
After almost a decade, I have made the decision to leave OpenAI. The company’s trajectory has been nothing short of miraculous, and I’m confident that OpenAI will build AGI that is both safe and beneficial under the leadership of @sama, @gdb, @miramurati and now, under the excellent research leadership of @merettm. It was an honor and a privilege to have worked together, and I will miss everyone dearly. So long, and thanks for everything. I am excited for what comes next — a project that is very personally meaningful to me about which I will share details in due time.
133
it's hilarious how the same people that refuse to engage with @ESYudkowsky's arguments are frothing at the mouth at the opportunity to use any dumb joke of his as a strawman to attack his real positions. Wonder why none of them never retweet the remaining 99% of his work...
I'm perturbed people fail to see how, not only does @ylecun's answer not address increased risk caused by future government military use of AGI. & infuriated that yet again an edge case is being to ignore the risk posed by all other possible bad actors (including the AI itself)
I’m perturbed people fail to see these obvious weak spots in doomer arguments.
& infuriated that yet again fear porn is being leveraged against the masses credulous nature .
1
192
how do I get a job with @Meta's alignment team? I have experience prompting GPT 3.5 with AI alignment homework.
2
158
1) His first company failed shortly after he was bailed out by a $43.4MM sale.
2) Now runs an incubator with around 10% success rate (and has made hundreds of millions from it).
@sama's reward function seems suboptimal for building tech that will kill everyone on first failure.
AI SAFETY SOLUTION: make every person working on models beyond GTP-4 have to survive 1 round of televised Russian Roulette.
If they are really ok risking a 25% chance of humanity dying in the name of progress, adding a measly 16.67% chance to their own deaths seems a fair ask.
There’s a ONE in FOUR chance AI causes human extinction - and AI companies admit it.
These same companies are trying to capture the world’s first AI Safety Summit to remain unrestricted and unregulated.
We’ve started this campaign to stop this from happening🧵👇
3
181
if only there was an easy way to test the "uterus hypothesis" without putting billions of women together on a train with unknown acceleration and an unknown upper speed limit...
oh wait, there is...
just one question, are you a ghostwriter for @pmarca's manifestos?
All new tech is scary, it's not just AI.
• trains — the female body was feared to not be able to go over 50 mph, lest the uterus "fly out of the body"
• subways — going underground was cursed because it brought you "closer to the devil"
• elevators — early riders got "elevator sickness" (physically ill) out of hysterical fear
Debatably, AI is the least scary invention on this list. The ones above all actually harmed people.
I fed @pmarca's manifesto as a prompt to Midjourney and got this image. Can someone explain?
unavailable
1
141
just came up with a new plan to slow down AGI: tell all the San Francisco homeless that there's a fentanyl lab hidden in the OpenAI office basement.
It's time to make our voices heard.
Reckless AI research programs led by double-digit p(doom)ers put our lives in jeopardy.
We will not accept their gamble.
We will not be the generation that erases the future. Join us to ⏸️:
Saturday, Oct 21 in SF: bit.ly/pause-ai-sf
1
1
143
I used to be terrified about AI x-risk. But after
@ylecun pointed out how the discovery of wooden clubs didn't cause human extinction, I can now sleep peacefully while my AI does my AI alignment homework.
6
165
Open sourcing LLMs with emergent properties we don't know of yet is such a great idea. Special thanks to @RichardSocher for helping me see how exciting this is.
1
87
Flawless logic by @bhorowitz, almost as good as his business partner @pmarca's AI-optimist arguments.
1
7
163
