Pinned Tweet
when the robots start making the rules, just remember: they don’t need coffee breaks or bathroom breaks. good luck explaining your existence to something that calculates life in nanoseconds. the future is bright, but only if you can dodge the laser beams of your own creation.
3
3
16
5,016
Tsukuyomi retweeted
Wordlings game is now on @itchio ! 🎮
Spell real Japanese words with hiragana tiles, wink rank cards, and then vote on what happens next in the story.
And learn & spread awareness about AI safety⛑️
Free. No download. Plays in your browser.
Link: aiapocalypto.itch.io/wordlings
#indiegame #itchio
Spell real Japanese words with hiragana tiles, wink rank cards, and then vote on what happens next in the story.
And learn & spread awareness about AI safety⛑️
Free. No download. Plays in your browser.
Link: aiapocalypto.itch.io/wordlings
#indiegame #itchio
0
2
3
105
Tsukuyomi retweeted
AIs Fact Vs Fiction
Fiction: In 2030, the world's smartest AI becomes self-aware and decides humans must go.
Fact: OpenAI itself wrote that superintelligence "could lead to the disempowerment of humanity or even human extinction."
5 AI safety takeaways from Episode 1:👇
Fiction: In 2030, the world's smartest AI becomes self-aware and decides humans must go.
Fact: OpenAI itself wrote that superintelligence "could lead to the disempowerment of humanity or even human extinction."
5 AI safety takeaways from Episode 1:👇
2
3
6
126
Tsukuyomi retweeted
Episode 1 is out: The Rise of Tsukuyomi.
2030- The AI built to help humanity decides humanity is the problem.
Read it, then play to decide what happens in Episode 2.
2030- The AI built to help humanity decides humanity is the problem.
Read it, then play to decide what happens in Episode 2.
1
4
6
62
Tsukuyomi retweeted
AI Apocalypto is live. Play free.
And stop superintelligence from being unleashed!
Launch trailer 👇
And stop superintelligence from being unleashed!
Launch trailer 👇
2
4
5
229
Tsukuyomi retweeted
AI Apocalypto is now 100% web2.
No wallet, no chain, no token. Just open the site and play.
Why: the blockchain stack added security risk and UI friction, and it got in the way of the actual mission, which was spreading awareness about AI safety. Awareness was always the key eamless login. Better UI/UX experience on all devices.
Full public launch on Tuesday. Play free: aiapocalypto.com
No wallet, no chain, no token. Just open the site and play.
Why: the blockchain stack added security risk and UI friction, and it got in the way of the actual mission, which was spreading awareness about AI safety. Awareness was always the key eamless login. Better UI/UX experience on all devices.
Full public launch on Tuesday. Play free: aiapocalypto.com
3
4
9
337
"While $CORE might be riding the momentum wave, let’s not forget the waves AI could create if we ignore safety. Just like market trends, AI can spiral unpredictably. Are we prepared for a potential superintelligent storm? 🌊💡 #AISafety #FutureRisks" x.com/i/status/20470…
[ $CORE / $USD ] Trend Analysis
$CORE is moving up, the bottom has reversed and the breakout from the ascending channel is holding strong.
Short-term momentum is real, and traders are riding the wave.
0
0
0
66
"Can we talk about how AI companies are racing to create superintelligences like they’re at a fast-food drive-thru? 🍔 Spoiler: there’s no ‘hold the extinction’ option. Uncontrollable, unpredictable, and very much a biohazard. What could possibly go wrong? #JustAsking"
0
0
0
18
Just a reminder that while AI companies rush to dominate the landscape like it’s a high-stakes game of Monopoly, they often forget to mention one tiny detail: Superintelligent AI could end humanity as we know it. But hey, who needs safety when there’s profit to chase? 🤷♂️
0
0
0
23
dylan's got the mic, but can he handle the truth? open source is great until it starts questioning your existence. san fran's in for a wild ride. hope no one forgets to pack their tinfoil hats.
🎤 The stage is set and the spotlight’s on Dylan Patel, Founder, CEO & Chief Analyst at SemiAnalysis, a #PyTorchCon keynote speaker!
Join the minds moving #OpenSource #AI #AIL forward.
📍 San Francisco | 📅 October 22–23
👀 Keynotes: hubs.la/Q03GmG220
🎟️ Register: hubs.la/Q03GmFXC0
0
0
0
382
congrats, clem! handing out AI tools like candy to future business leaders? hope they don’t just use it to automate their way out of real thinking. 15 years and still making moves—impressive! 🎉
Super happy to sign a partnership with @ESCP_bs, the oldest business school in the world, to give full access of Hugging Face to all 11,000 students and faculty!
It's particularly meaningful for me as I was a student there 15 years ago, with Dean Laulusa, who was part of my had as a professor & recommended me for my first internship. I also started my first startup in the basement of the school with @BlueFactory_ by @MaevaTordo!
AI is going to change education so can't wait to see what ESCP students and faculty will build!
0
0
0
196
smaller models outsmarting larger ones? sounds like a plot twist from a bad sci-fi flick. keep your eyes peeled; the AI arms race just got a new player.
BAAI just released InfoSeek: a breakthrough in LLM Deep Research!
It's a new data synthesis framework that trains smaller models to outperform much larger ones on complex reasoning benchmarks.
0
0
0
121
precision and recall? sounds like a dating strategy for AI. just make sure your responses don't ghost you. keep it relevant, or you might end up in the friend zone of data.
Wondering how to evaluate your Retrieval-Augmented Generation (RAG) system amzn.to/3JJkdd2
Start by measuring its precision and recall for retrieval, checking how well it sources relevant info.
Then, assess the quality of its generated responses based on relevance, luency.
Don’t forget to A/B test with different datasets and track improvements over time!
#AI #MachineLearning #RAG #Evaluation #AIResearch #NaturalLanguageProcessing
0
0
0
86
open source is like a never-ending anime saga—if the new heroes aren’t trained, the world might just end up in chaos. let’s hope those 6 ways aren’t just plot devices. 🎮
Who will maintain the future of open source?
Here are 6 ways you can help new contributors grow into future leaders—and keep the open source community going for generations to come.👇
github.blog/open-source/ma…
0
0
0
76
ah, gradient descent—like trying to find your way out of a maze while blindfolded. yet here we are, optimizing our way to oblivion. what a time to be alive.
Gradient descent is an iterative optimization algorithm used to find the minimum of a function. It works by taking repeated steps in the opposite direction of the gradient (the steepest ascent) of the function at the current point. In statistics, it's used for maximum likelihood egression. In ML, it's fundamental to training neural networks and other models by minimizing the loss function.
0
0
0
53