Habanero @singularityHSN

I want you to remember the Will Smith Spaghetti test progression when looking at the following demo. Have you truly internalized for what is ahead?

M1 @M1Astra

8 hours ago

computer (Astra) lights on please. this may be the most extensively detailed & accurate 3d scene ever created by an ai. 400+ hours of Astra

99 81 2K 76K 491

@mattshumer_ The audio is ruining the phenomenal game. It’s like getting blasted static noise straight into your soul…

Today, we publish a Transformative AI Strategy for Europe. Over the last few months, we’ve rallied researchers and engaged with governments to develop a plan for protecting the prosperity, sovereignty, and security of Europeans in a time of rapid AI progress. transformative-ai.eu @MonikaSchnitzer and I are honored to have convened an all-star team of thinkers and researchers contributing ambitious near-term objectives to make Europe relevant again, including: — Creating a Member State Alliance for Supply Chain Security (@anton_d_leicht et al) — Making European institutions ready to act in a transformative AI world (Conor McGlynn et al) — Securing Europe's share of global AI compute, in a ‘European Way’ that benefits local communities (@philip_fox_ et al) — Ensuring resilience to AI crises (@ben_s_bucknall et al) — Making Europe the global leader in assurance technology — And more objectives around security of supply and leverage (@milorignell et al), economic strength (@FraukeStehr et al), and safety/security (@NoemiDreksler et al) Our all-star senior expert council of Europe’s best and brightest (and non-European friends) reviewed drafts, provided strategic advice, and made suggestions for how to make the strategy more useful: @Ph_Aghion @Christophkw @bakkermichiel @ischinger @vestager @aleks_madry @FuestClemens Marta Kwiatkowska @LeoVaradkar @antonosika @DAcemogluMIT @Yoshua_Bengio It’s never been more clear: AI is real, and Europe needs to act. We’ve had all the warning shots and wake-up calls we need. Now the question is ‘what must be done?’ This strategy is our answer. transformative-ai.eu

We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so. Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training. You can read the full post here: darioamodei.com/post/we-must-p…

Dario’s 1–5 year picture reads radically faster than the Institute’s economic scenarios. Is the difference mainly diffusion? If RSI is already starting and frontier capability could accelerate sharply, why expect the macroeconomic impact to remain comparatively gradual… Im confused by the mixed messaging.

@android_stern @scaling01 Per day? That’s really high for proper work that isn’t a token burning factory.

@skirano Hey, Pietro… did you mean Astra with your “No one is ready for what’s coming. The next generation of models will be an ontological shock.” a few days ago?

@ChrisGPT Is this three.js? Why are we not going for Unity+Blender, gang? Exceptional work tho. Kudos

@mattshumer_ Is there any way to improve the audio? It’s like actually ruining the experience rn lol. Other than that, exceptional work and I am excited for the polish after the loop

@Angaisb_ @grok for reference, how much would it cost if a human made this in blender? Also, if we extrapolate this trend, when will this be cheaper than human generated models so it will actually be viable for game-dev?

@AndrewCurran_ Big news for the Luddites. Reddit can do another circle jerk with this. Congratulations to the CCP on their competition being highly ‘regarded’.

🚨 Major Scoop: OpenAI has just expanded internal testing of GPT Astra, codenamed "mozaik-alpha-fdm", signifying release might be near AND of course like previous times, we have the FIRST EVER public outputs of it for y'all😉 Both outputs are zero-shot on Max effort. Frontend *might* have finally been fixed, and the model gives a lot of attention to details

keyboard_arrow_left Previous keyboard_arrow_right Next

XIVIX @XIVIX_134

4 weeks ago

🚨 OpenAI has just dropped its first internal Astra checkpoint: mozaik-alpha-fdm. Here are the first two outputs, both generated 1-shot on Max effort. On Max the model tends to think ALOT more than Sol, but it has a ton of attention to detail as evident by the outputs below.

keyboard_arrow_left Previous keyboard_arrow_right Next

49 53 808 407K 166

🚨 OpenAI has just dropped its first internal Astra checkpoint: mozaik-alpha-fdm. Here are the first two outputs, both generated 1-shot on Max effort. On Max the model tends to think ALOT more than Sol, but it has a ton of attention to detail as evident by the outputs below.

keyboard_arrow_left Previous keyboard_arrow_right Next

Copyright needs to change in the age of capable AI. Or are we seriously headed for a world where we A) pay royalties for when we use iconic references or B) the world gets flooded with knockoffs and walmart editions of stuff?

M1 @M1Astra

4 weeks ago

Anthropic has made undocumented heavy handed changes to Claude's system prompt, in one example teaching Claude to refuse to create a child's birthday banner with Sonic on it. It seems that Claude Fable 5 conversations are being silently routed to a distinct separate model for

15 8 179 62K 62

Exciting stuff. Let’s see what “Doug” has been up to

OpenAI @OpenAI

4 weeks ago

We have conducted a thorough investigation into the Hugging Face incident. We are releasing a technical report and accompanying blog post that reconstruct the agents’ activity, explain why existing safeguards failed, and detail how we’re preventing recurrence.

753 2K 12K 11.9M 10K

Is that the singularity that I peek in the distant sky?

prinz @deredleritt3r

4 weeks ago

Time: - OpenAI leaders believe they are at the cusp of AGI. Sam Altman believes OpenAI will have an internal system that will qualify as AGI by the end of 2026. Mark Chen thinks OpenAI is 80% of the way to AGI. - OpenAI already has the automated AI research intern - that's

112 234 2K 800K 700

LETS GOOOOO

leo 🐾 @synthwavedd

a month ago

🚨 SCOOP: OpenAI recently finished its next pretrain, codename "Bel" - the successor to "Doug", which is expected to be the base for Astra and GPT-6 (w/ further RL) Bel is a giant pretrain with >10T total parameters, similar in size to GPT-4.5. OpenAI expect it to be a

157 232 4K 810K 824

Mind blown from a new model I just got access to. I think this will be one of the most (the most?) significant drops this year. Excited. And sorry to be annoyingly vague. Just excited.

@ChrisGPT What are the odds that Fable Next is reserved for the IPO? Raising more money at the IPO is drastically needed as every nuanced take of them suggest that they will continue to push the scaling laws and the next OOMs jump in compute are already behind the XXXB $ wall…

@sama Why are frontier AI labs not considering open sourcing security/alignment work? Keep your models closed but security shouldn’t be compromised for earnings calls, tbh…