Co-Founder / CTO @ClipStake_X | Backed by @Alliance
New York, NY
Joined April 2015
- Tweets485
- Following700
- Followers547
- Likes456
Pinned Tweet
Slowing down AI development won’t be effective. The return on investment of developing a model that is close to the frontier model far exceeds the research and training costs.
0-days, breakthrough scientific research, new drug discovery, …
Over time, the cost of a unit of intelligence will decrease significantly. If it becomes unavailable to the public, open-source development will take over.
Supply and demand.
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…
alex retweeted
Watch Starship Flight 14 xitter.cf/i/broadcasts/1qxvvenMP…
Now add artificial intelligent players playing amongst themselves trying to achieve specific goals, and you have an mvp of our reality/world
Demand pressure for magnets is increasing significantly; Actuators. Companies are now paying a high premium to lock in future supply.
Similar to how NVIDIA locked in fab capacity early on.
Part of @stripe MPP and x402 strategy, and @a16z made around $1.2B profit on this acquisition (~60x their initial $20M investment / owning 17% of shares)
OpenRouter is joining Stripe.
We started OpenRouter with a simple mission: intelligence should be multi-model. Today, we are the largest AI marketplace & gateway, processing 10T+ tokens daily on 400+ models.
Joining @Stripe gives us the opportunity to accelerate that mission.
The trick is a lightweight "selector" that traces one coherent path through the draft model's parallel token candidates, plus a two-tap convolution that stops draft quality from decaying near the end of each block. Never seen that before, pretty cool
DFlash 2 is here! Qwen3.8-27B at 70 tok/s on an M5 Max MacBook Pro.
⚡ Up to 4.6× the speed of autoregressive decoding, with the same output.
This is the next generation of DFlash, seeded at Z Lab and upgraded at Inco AI. Get one more accepted token on every pass, for free!
inco.ai/blog/dflash2/
alex retweeted
We just shipped our official Qwen 3.8 27B Uncensored MLX build. Local. Uncensored. For🍎
2-bit, 4-bit, 6-bit & 8-bit — pick your poison based on RAM and speed.
No CUDA. No cloud. Just your Mac and the weights. Have fun!
huggingface.co/orcarouter/Qw…
The main reason X was open-sourced @XOpenSource is to expose how governments are control freaks and corrupt.
Governments send countless letters with takedown and censorship requests to all big social media platforms, all under NDAs.
Platforms are forced to comply; otherwise they get taken down. Countless examples have already happened:
- Nepal, 2025: Blocked Facebook, Instagram, WhatsApp, YouTube and X for failing registration requirements.
- Brazil, 2024: Blocked X after it failed to comply with court orders.
- Turkey, 2024: Blocked Instagram during a dispute over content moderation and compliance with Turkish law.
- Albania, 2025: Blocked TikTok over youth safety and harmful-content concerns.
- Pakistan, 2024: Blocked X amid political unrest and disputes over platform access.
- France, 2024: Arrested Telegram founder Pavel Durov over alleged failures to cooperate with authorities and police illegal activity on Telegram.
Every government tries to control big (social) media.
@X wants to be the truth-seeking platform without censorship. The only way is to force governments to expose themselves and play by the same rules:
We’re open source; whatever you impose on us will be visible to the world.
Genius.
Today’s For You release is in the open-source repo.
Notable updates:
• How ranking weights actually work
• A new filter required for Brazil’s 2026 election
All the details: github.com/xai-org/x-algorit…
Funny how @Kimi_Moonshot cut 20% of their usage plans released in the first days after Kimi3 dropped, and no one noticed. They managed to fill new spots from their waitlist like this.
alex retweeted
Hit your usage limit in Claude Code desktop? There's now an auto-continue checkbox.
Turn it on, and it'll automatically continue where you left off once your limit resets.
alex retweeted
Everyone on my timeline was losing it this week over @emergentlabs raising $130M at a $1.5B valuation.
So I gave the same prompt to @boltdotnew @v0 @lovable_dev and Emergent, and read every line of code.
One of them shipped a 2019 stack in 2026. Create React App. CRACO. Mongo. No TypeScript.
Full teardown 👇
▶️ recommended watch speed: 1.5x
00:00 The $1.5B question
00:50 The prompt — same for all 4
01:15 @boltdotnew goes first, no pressure
08:48 @emergentlabs enters to a standing ovation
09:41 the UI is a genuine 10/10
12:11 …then I opened the code
14:30 Create React App + CRACO in 2026
15:08 @theo called this exact thing 4 years ago
19:33 Mongo, for a relational app (@theo again)
21:00 the Emergent scorecard
22:23 @v0 has thoughts on your design taste
28:06 @lovable_dev, living up to the name (mostly)
35:03 who I'd actually ship to production
35:50 my questions for @emergentlabs + YC
Both @theo videos are linked below — genuinely worth the watch 🧵