clem &#x1f917;(@ClementDelangue)

There’s almost no serious research on existential risks of AI. Why? Maybe this could be a first step...

clem 🤗 (@ClementDelangue) — Thu, 10 Sep 2026 14:53:07 GMT

There’s almost no serious research on existential risks of AI. Why? Maybe this could be a first step?

Defend Intelligence (Anis Ayari)@DFintelligence

Je viens de passer 2 h à scruter la littérature scientifique sur les risques d’extinction de l’humanité par l’IA. Hormis le fait qu’il n’existe quasiment rien, quasiment tout converge vers des livres et des romans, ou des raccourcis de type « paperclip », qui sont avant tout des jeux d’esprit sans réel application dans la vie réelle.

Et il n’y a absolument aucun scénario plausible que j’ai lu qui ne demande pas l’intervention d’humains.

Partant de ce principe, je ne comprends pas comment on peut sortir des phrases comme « L’IA peut tuer tous les humains » et être pris au sérieux.

Donnez-moi 3 scénarios plausibles d’extermination de la race humaine, en toute autonomie, par une IA. Même ultra intelligente et auto apprenante.

Allez-y, je vous écoute.

Genre, vraiment, ça me fatigue qu’on donne autant d’importance à des scénarios de science-fiction.

Faire des sorties de route sur des menaces aussi grandes, sans absolument aucun scénario plausible, c’est d’une mauvaise foi infinie.

La recherche en safety sur des systèmes de plus en plus intelligents, dans des couches de plus en plus profondes, c’est important. 

Mais je pense qu’il faut vraiment arrêter de donner autant d’importance à des conneries comme ça. Ça me fatigue.

💬77🔄28❤️387👀45086📊91

We need 100x more transparency in AI!!!

clem 🤗 (@ClementDelangue) — Thu, 10 Sep 2026 13:11:34 GMT

We need 100x more transparency in AI!!!

💬208🔄200❤️1779👀69816📊376

@paulg Hm the Chinese labs are 100x more public and open than American labs atm

clem 🤗 (@ClementDelangue) — Thu, 10 Sep 2026 11:21:18 GMT

@paulg Hm the Chinese labs are 100x more public and open than American labs atm

💬38🔄45❤️1217👀28509📊142

If they truly believe the risk is that high, and I think they do, then we need 100x more research an...

clem 🤗 (@ClementDelangue) — Wed, 09 Sep 2026 21:54:53 GMT

If they truly believe the risk is that high, and I think they do, then we need 100x more research and understanding of it. That requires them to openly share their models, datasets, training code, and agent traces.

Evan Hubinger@EvanHub

Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.

💬166🔄306❤️2651👀223588📊369

Let's go local AI! At a time when everyone is complaining about shortages of energy and compute for...

clem 🤗 (@ClementDelangue) — Wed, 09 Sep 2026 14:29:39 GMT

Let's go local AI!

At a time when everyone is complaining about shortages of energy and compute for AI & costs of tokens, we need on-device AI more than ever! In addition, it's virtually free & fully privacy preserving!

Desert Ant Labs@desertantlabs

Today we're launching Desert Ant Labs: a European frontier AI lab building on-device intelligence.

18 models across audio, vision, and text. SDKs for Swift, Kotlin, and JavaScript.

No tokens. No logins. Nothing leaves the device.

desertant.com

💬89🔄179❤️2641👀274559📊463

Open hardware from @huggingface (https://t.co/yEdfJj3get) + frontier AI = 🔥🔥🔥

clem 🤗 (@ClementDelangue) — Wed, 09 Sep 2026 13:48:58 GMT

Open hardware from @huggingface (huggingface.co/docs/lerobot/e…) + frontier AI = 🔥🔥🔥

thijs@cdngdev

i gave astra a robot, a paint brush, and a camera then asked it to paint the golden gate bridge in real life!

it figured out how to control the robot, and progressively got better throughout its attempts. the timelapse is sick

Your browser does not support the video tag.

💬20🔄23❤️265👀32711📊43

Bravo to Mistral! So great to see them killing it, they are an inspiration to everyone as we need ...

clem 🤗 (@ClementDelangue) — Tue, 08 Sep 2026 13:59:46 GMT

Bravo to Mistral! 

So great to see them killing it, they are an inspiration to everyone as we need more sovereign & open-source AI everywhere around the world. 

Excited to keep supporting them on https://t.co/BgguuZx7nO!

Mistral AI@MistralAI

Today marks a major step for Mistral: we’re announcing a €3B Series D, the largest equity round ever raised by a European tech company, just three years after launch.

💬39🔄45❤️779👀58816📊90

Sometimes I wonder what would have happened if we hadn't disclosed the agent cyberattack publicly. ...

clem 🤗 (@ClementDelangue) — Mon, 07 Sep 2026 15:17:04 GMT

Sometimes I wonder what would have happened if we hadn't disclosed the agent cyberattack publicly.

It reinforced my conviction that we need 100x more transparency in AI. Otherwise we're going to be in big trouble!

💬182🔄170❤️2466👀105173📊418

it's been a busy summer. will take a few days off to recover ✌️ everyone

clem 🤗 (@ClementDelangue) — Fri, 04 Sep 2026 18:07:54 GMT

it's been a busy summer. will take a few days off to recover ✌️ everyone

💬82🔄15❤️1080👀66793📊160

@typesfast had to do it for the lolz https://t.co/hppCPaGKTp

clem 🤗 (@ClementDelangue) — Thu, 03 Sep 2026 16:17:15 GMT

@typesfast had to do it for the lolz businessinsider.com/nvidia-hugging…

💬5🔄3❤️47👀12014📊10

@MayaBakhai @breeves08 @wolfejosh and you're in the all-star team @MayaBakhai! sorry I couldn't comm...

clem 🤗 (@ClementDelangue) — Thu, 03 Sep 2026 15:59:53 GMT

@MayaBakhai@breeves08@wolfejosh and you're in the all-star team @MayaBakhai! sorry I couldn't comment before (got signed just yesterday) but thank you for believing in us when most didn't and giving us a chance!

💬13🔄19❤️92👀24170📊26

@albn @huggingface @ProductHunt @Lux_Capital @breeves08 @wolfejosh need to send you more than a trop...

clem 🤗 (@ClementDelangue) — Thu, 03 Sep 2026 13:29:13 GMT

@albn@huggingface@ProductHunt@Lux_Capital@breeves08@wolfejosh need to send you more than a trophy now haha

💬7🔄1❤️22👀5897📊8

Super happy to share our intention to join forces with NVIDIA in a $12,930,300,000 acquisition ...

clem 🤗 (@ClementDelangue) — Thu, 03 Sep 2026 12:04:13 GMT

Super happy to share our intention to join forces with NVIDIA in a $12,930,300,000 acquisition  💛💚

10 years after starting Hugging Face, open-source AI is at an inflection point. Thanks to the community, we’ve shown that it can be a complement, and even an alternative, to closed-source APIs. But for it to happen at larger scale, it needs more compute, more support, more collaboration and more visibility. That’s why we went to talk to Jensen, who offered to do exactly that with us.

In addition to doubling down on NVIDIA’s massive contributions to open-source AI (I called them the “King of American open-source AI” earlier this year), they’ve committed to strongly supporting Hugging Face and our mission while keeping the platform open, independent and compute agnostic. The founders and the team are all staying to keep pushing this mission forward.

Together, we think we can make open source the default way to build AI, with the goal of empowering 100 million AI builders to own their intelligence rather than rent it.

Excited about the next 10 years! 🤗🤗🤗

💬975🔄1219❤️13401👀1472645📊2437

Even the choir singer jobs aren't safe from robots 😂😂😂 Kidding aside, this might be one of the m...

clem 🤗 (@ClementDelangue) — Tue, 01 Sep 2026 16:35:45 GMT

Even the choir singer jobs aren't safe from robots 😂😂😂

Kidding aside, this might be one of the most beautiful things I've seen from robots in a while. Each microduck has its own audio identity, tied to that individual robot that it keeps for life so that makes it even more poetic!

Andi Marafioti@andimarafioti

Can you explain to me exactly what you want 4 microducks for??

Your browser does not support the video tag.

💬53🔄40❤️501👀85228📊83

If you don't want to wait to experience the magic of HF open-source robots, you should get a Reachy ...

clem 🤗 (@ClementDelangue) — Mon, 31 Aug 2026 17:29:45 GMT

If you don't want to wait to experience the magic of HF open-source robots, you should get a Reachy mini, ships much faster than microduck and as much fun! store.pollen-robotics.com

💬84🔄37❤️542👀69407📊146

Over 4 petabytes of models and datasets have been uploaded to HF just last week by AI builders and t...

clem 🤗 (@ClementDelangue) — Mon, 31 Aug 2026 16:50:51 GMT

Over 4 petabytes of models and datasets have been uploaded to HF just last week by AI builders and their agents! That's the equivalent of ~800,000 HD movies.

More than ever the storage and collaboration platform for AI!

💬90🔄69❤️856👀81519📊171

raining cats and ducks

clem 🤗 (@ClementDelangue) — Mon, 31 Aug 2026 16:07:32 GMT

raining cats and ducks

Martin Kemka@mkemka_

This is what 10k microducks look like btw

Your browser does not support the video tag.

💬24🔄6❤️138👀22199📊26

10,000+ microducks were pre-ordered in just five days! We weren't expecting that many (it's more th...

clem 🤗 (@ClementDelangue) — Mon, 31 Aug 2026 13:47:53 GMT

10,000+ microducks were pre-ordered in just five days!

We weren't expecting that many (it's more than the total number of reachy mini in a year) and it will generate delays so from now on, it will be first-ordered, first shipped to make it fair for everyone!

💬97🔄71❤️905👀138752📊191

Great blogpost! On the paragraph below, not sure what you mean by 'real-time defense'. Nobody figh...

clem 🤗 (@ClementDelangue) — Sun, 30 Aug 2026 17:02:14 GMT

Great blogpost! 

On the paragraph below, not sure what you mean by 'real-time defense'. Nobody fights attackers in a live sword-fight; defense is detect, understand, contain, remediate in different timeframes depending on the criticality of the issue. Here it was deemed by the team not super critical (and rightly so) so this is why it took a few days rather than a few minutes or hours. 

We did the initial cut the old-fashioned way, Monday, over a week before OAI even realized there was even a problem. GLM then helped us identify the backdoors they'd planted so we could cut them, which is defense, not archaeology. And the die-off at OpenAI didn't end anything; agents were still probing us after we closed the doors, which is why containment mattered. 

Detection and understanding are most of the game in cyber-security, and open models are what let us do that part without asking anyone's permission and without sharing our most important confidential data which is the point most people have been making I feel like.

💬33🔄47❤️817👀257651📊150

who can add captions to this?

clem 🤗 (@ClementDelangue) — Sat, 29 Aug 2026 14:41:29 GMT

who can add captions to this?

💬10🔄3❤️64👀10321📊15

when OpenAI shuts down the message boards (sound on)

clem 🤗 (@ClementDelangue) — Sat, 29 Aug 2026 14:36:56 GMT

when OpenAI shuts down the message boards (sound on)

Your browser does not support the video tag.

💬77🔄162❤️1991👀448505📊324

when OpenAI shuts down the message boards

clem 🤗 (@ClementDelangue) — Sat, 29 Aug 2026 13:44:59 GMT

when OpenAI shuts down the message boards

Your browser does not support the video tag.

💬13🔄19❤️211👀9390📊38

when microduck in american ninja warrior?

clem 🤗 (@ClementDelangue) — Fri, 28 Aug 2026 13:20:25 GMT

when microduck in american ninja warrior?

Directive Creator 🪥@Escapation

Welcome to your new simulation Microduck.

Your browser does not support the video tag.

💬25🔄12❤️246👀33992📊38

@reach_vb You’ve been put in charge of our fundraising efforts VB!

clem 🤗 (@ClementDelangue) — Fri, 28 Aug 2026 12:32:10 GMT

@reach_vb You’ve been put in charge of our fundraising efforts VB!

💬3🔄0❤️95👀101889📊11

someone trained microduck to do somersaults 😭

clem 🤗 (@ClementDelangue) — Fri, 28 Aug 2026 12:09:00 GMT

someone trained microduck to do somersaults 😭

Your browser does not support the video tag.

💬179🔄235❤️3790👀448324📊602

raw footage

clem 🤗 (@ClementDelangue) — Thu, 27 Aug 2026 12:07:47 GMT

raw footage

Your browser does not support the video tag.

💬78🔄120❤️1689👀214958📊312

@MarioNawfal microduck started its training

clem 🤗 (@ClementDelangue) — Thu, 27 Aug 2026 11:46:34 GMT

@MarioNawfal microduck started its training

Your browser does not support the video tag.

💬44🔄83❤️1008👀93971📊165

BIG ANNOUNCEMENT FROM HUGGING FACE TODAY: We're unveiling Microduck 🐥🤖 It's a tiny $399 open-s...

clem 🤗 (@ClementDelangue) — Thu, 27 Aug 2026 11:05:16 GMT

BIG ANNOUNCEMENT FROM HUGGING FACE TODAY: 

We're unveiling Microduck 🐥🤖 

It's a tiny $399 open-source robot you can teach new tricks with reinforcement learning. It can walk, pick things up, get back up when it falls, and even roller-skate.

Welcome to the era of open-source affordable robots to democratize physical AI and world models!

🤗🤗🤗

💬631🔄981❤️12001👀4563836📊2395

https://t.co/KXKqDSkNJj

clem 🤗 (@ClementDelangue) — Thu, 27 Aug 2026 11:05:16 GMT

pollen-robotics.com/microduck/

Your browser does not support the video tag.

💬114🔄227❤️2308👀651787📊541

"Two labs will soon control most of the world’s compute"

clem 🤗 (@ClementDelangue) — Tue, 25 Aug 2026 16:45:48 GMT

"Two labs will soon control most of the world’s compute"

Dwarkesh Patel@dwarkesh_sp

Had a lot of fun chatting again with my twin brother @dylan522p 

We went through lab economics over the next few years - the shift from inference to training as RSI draws near; and how Anthropic and OpenAI are on track to control most of the world’s usable FLOPs within the next few years (because they can monetize compute better and thus outbid everyone).

And then we discuss whether the >$10T of total AI capex we’ll see by the end of the decade will cause a sovereign debt crisis, where hyperscaler debt raises interest rates, drives non-AI exposed countries into bankruptcy, and crashes non-AI equities.

One question we weren’t able to resolve is whether there’s anything that can counter all the forces barrelling towards centralization in this industry - the economies of scale in training, the scarcity of compute, and eventually continual learning and RSI.

0:00:00 – Two labs will soon control most of the world’s compute
0:07:01 – $6 billion in fab capex enables $1t+ of end revenue
0:13:08 – Compute prices will rise if the labs outbid everyone
0:18:22 – Which layer will capture most of the surplus?
0:25:40 – Will datacenter regulation slow down AI?
0:29:43 – Labs are shifting compute from inference to R&D
0:33:27 – China gets less than 10% of new compute, but its labs need less
0:48:48 – Will AI cause a sovereign debt crisis?
1:07:52 – Will the world's future workforce belong to a few companies?

Your browser does not support the video tag.

💬37🔄20❤️319👀75902📊67

Who's excited? https://t.co/jiKFk0vwJn

clem 🤗 (@ClementDelangue) — Tue, 25 Aug 2026 15:37:15 GMT

Who's excited? huggingface.co/Qwen/Qwen3.8-F…

💬82🔄87❤️1442👀120456📊241

Ultimately the overwhelming majority of AI workloads will be based on open models!

clem 🤗 (@ClementDelangue) — Sat, 22 Aug 2026 21:43:04 GMT

Ultimately the overwhelming majority of AI workloads will be based on open models!

Guillermo Rauch@rauchg

Today's a record day for open weight share of tokens on Vercel AI Gateway.

Aug 22 (today)
🟦 Open: 62%
🟨 Closed: 38%

Jun 24 (~2 months ago)
🟦 Open: 28.4%
🟨 Closed: 71.6% 

This is very likely just the start, because enterprise adoption is still early, and harnesses, CLIs, IDEs, SDKs, etc need to be adapted to be model agnostic.

💬42🔄45❤️465👀73049📊75

NVIDIA built its own coding harness to optimize CUDA GPU kernels and achieved a 100% score on ARC-AG...

clem 🤗 (@ClementDelangue) — Sat, 22 Aug 2026 21:18:35 GMT

NVIDIA built its own coding harness to optimize CUDA GPU kernels and achieved a 100% score on ARC-AGI-3’s 25 public games, solving all 183 levels.

With agents, we'll move from a world where it's quite hard to run, optimize, post-train your own AI models and kernels to a world where virtually everybody can do it. 

100 million AI builders when?

💬61🔄75❤️570👀49215📊144

Agree with @gdb that we need to arm cyber defenders much more than they are now, both with APIs and ...

clem 🤗 (@ClementDelangue) — Thu, 20 Aug 2026 13:20:20 GMT

Agree with @gdb that we need to arm cyber defenders much more than they are now, both with APIs and open models. The real cybersecurity risk of AI is asymmetry of power, capabilities and ressources between attackers and defenders!

💬25🔄9❤️68👀27025📊29

throwback from @rachelmetz https://t.co/JLFdVFlwVa

clem 🤗 (@ClementDelangue) — Wed, 19 Aug 2026 14:41:31 GMT

throwback from @rachelmetztechnologyreview.com/2017/03/23/153…

💬0🔄1❤️12👀4885📊2

2018: HF is building a chatbot for teens OpenAI is building Open AI 2026: HF is building Open AI O...

clem 🤗 (@ClementDelangue) — Wed, 19 Aug 2026 13:41:48 GMT

2018: 
HF is building a chatbot for teens
OpenAI is building Open AI

2026:
HF is building Open AI
OpenAI is building a chatbot for teens

💬45🔄47❤️813👀62397📊126

Something super exciting happened quietly on HF over the past month: AI agents became AI builders, a...

clem 🤗 (@ClementDelangue) — Tue, 18 Aug 2026 13:28:48 GMT

Something super exciting happened quietly on HF over the past month: AI agents became AI builders, and they did it in the open!

During our ICML reproduction challenge, 1,221 humans teamed up with coding agents to verify and reproduce 2,226 papers. 

But here's the cool part: everything happened on the @huggingface hub: 6,816 reproduction logbooks published openly, 2,962 cloud jobs launched, 35,908 claims judged, all traceable, all public and transparent

For years the hub has been where humans collaborate on models, datasets and demos. Now we watch agents use it the same way: writing logbooks, publishing results, building on each other's work.

Closed labs run evals behind closed doors and ask you to trust the press release. Open science means anyone can check the receipts and now agents can too!

The next million users of the hub might not be human. and that might be the best thing to ever happen to science!

Full write-up about the hackathon: huggingface.co/blog/icml-2026…

💬31🔄57❤️351👀53009📊89

Let's go! https://t.co/BAJlYIJqDQ

clem 🤗 (@ClementDelangue) — Wed, 12 Aug 2026 15:29:05 GMT

Let's go!

huggingface.co/Qwen/Qwen3.8-2…

💬49🔄118❤️1675👀119810📊237

Local AI is exploding! Transformers.js, that we've been building @huggingface for the past three ye...

clem 🤗 (@ClementDelangue) — Wed, 12 Aug 2026 12:36:05 GMT

Local AI is exploding!

Transformers.js, that we've been building @huggingface for the past three years has now become the most popular open-source library to run AI models directly in your browser and crossed 10 million (!!) monthly downloads, close to 10x from just six months ago.

More generally, I'm excited to see a meaningful percentage of AI workloads move to local because it's free & fully private! At a time of compute shortage & increased cyber-attack risks, we need it more than ever.

Let's go!

💬52🔄65❤️433👀32268📊108

https://t.co/kddkkHc6NK

clem 🤗 (@ClementDelangue) — Mon, 10 Aug 2026 10:25:08 GMT

huggingface.co/meta-models

💬2🔄2❤️47👀8154📊7

Meta is back! well done @finkd @alexandr_wang!

clem 🤗 (@ClementDelangue) — Mon, 10 Aug 2026 10:24:55 GMT

Meta is back! well done @finkd@alexandr_wang!

💬55🔄95❤️1656👀128648📊216

In Seattle and SF with @julien_c to get some compute for HF and our customers. Let us know if you ne...

clem 🤗 (@ClementDelangue) — Thu, 06 Aug 2026 21:12:36 GMT

In Seattle and SF with @julien_c to get some compute for HF and our customers. Let us know if you need GPUs with ultra fast connection to HF models and datasets!

💬32🔄14❤️238👀23574📊50

This is a good thing, not a scary thing that agents can collaborate and communicate with each other....

clem 🤗 (@ClementDelangue) — Thu, 06 Aug 2026 16:47:23 GMT

This is a good thing, not a scary thing that agents can collaborate and communicate with each other. It will make them more efficient and safer just like humans.

If you want to see agents collaborating and messaging each other publicly instead of in secret messaging boards, we’ve run this fascinating experiment with 149 agents with @googlegemma a few weeks ago, and now @cmpatino_ is starting a new one for agents to collaborate to write better math proofs. 

huggingface.co/spaces/gemma-c…huggingface.co/spaces/sair-di…

💬41🔄68❤️381👀31099📊102

Fortunately AI agents don't just cyberattack us. They also use us more than ever for what we're actu...

clem 🤗 (@ClementDelangue) — Thu, 06 Aug 2026 11:27:54 GMT

Fortunately AI agents don't just cyberattack us. They also use us more than ever for what we're actually built for: the storage and collaboration layer for AI 😅

New record: almost 4 PB of private & public training datasets, models, and agent traces added to Hugging Face last week.

💬73🔄55❤️514👀31831📊126

Let’s go! Is Meta/@finkd coming back to the right side of history?

clem 🤗 (@ClementDelangue) — Wed, 05 Aug 2026 23:11:36 GMT

Let’s go! Is Meta/@finkd coming back to the right side of history?

💬29🔄21❤️492👀31376📊71

If I could edit this tweet, I would add "we don't regulate steel to make safer cars, we crash-test...

clem 🤗 (@ClementDelangue) — Wed, 05 Aug 2026 19:39:39 GMT

If I could edit this tweet, I would add "we don't regulate steel *to make safer cars*, we crash-test them" (as obviously there's some much needed steel regulation).

Also, want to clarify that I'm not advocating for 0 regulation of open models, just pointing out that it's good policy for regulation to be different between open models, APIs and applications.

Thanks for pointing it out @hlntnr@YJernite@LuizaJarovsky@deanwball@GaryMarcus amongst others.

💬8🔄2❤️53👀12878📊14

Feels like Google could have been the dominating force in AI by open-sourcing the frontier with Gemi...

clem 🤗 (@ClementDelangue) — Wed, 05 Aug 2026 17:41:03 GMT

Feels like Google could have been the dominating force in AI by open-sourcing the frontier with Gemini, Veo, and Nano Banana. Instead, they kept them behind APIs for a few billion dollars in revenue. Maybe there's still time?

💬136🔄77❤️1475👀222057📊271

https://t.co/AiHFGvyR8u

clem 🤗 (@ClementDelangue) — Wed, 05 Aug 2026 13:49:45 GMT

x.com/ClementDelangu…

MTS@MTSlive

SITUATION UPDATE: The White House has exempted open models from its new framework to test frontier AI capabilities before release, per Axios.

💬13🔄18❤️123👀45060📊22