Head of Foundation Models @Salesforce 🤖 Founder @SpindleAI (acquired @Salesforce) 🛠 ex-President Brainspace AI (acquired, Reveal) ex-CTO Volley ML (@JPMorgan)

SF
Joined April 2010
Introducing SubQ Snap, the first long-context decision model for real enterprise work. Snap is the only decision model or LLM where compute per token stays constant as context grows. • 384K tokens of multimodal context, 12x larger than Jev. • 61.7 on Decision Index v0.2.1 vs Jev’s 57.9, outperforming on 13 of 16 enterprise benchmarks. •Processes 100K tokens and returns a decision in under one second. •$0.05/MM input tokens. Output is free. Available today to anyone. The first 10,000 sign-ups get 100M free input tokens. We’ll likely hit capacity fast and will add users as we add compute. Compute offers welcome 🙂. Latency will be variable as we tune the platform to this new model and category of users. Model speeds are 80ms for minimal context, 250ms for 32k, 660ms for 100k, and 2.5 seconds for 384k. Visit: subq.ai/
575
647
267
3,069
4,500,678
Training a model to generate RL tasks not too hard, not too easy costs many solver runs per task. PROPEL predicts difficulty via a probe on its activations instead, amortizing cost and speeding up generator optimization. New open-ended RL research from @Vmax + @GoodfireAI.
7
26
10
135
32,604
Carson Kahn retweeted
Wrote a short blog about the "shape" of language models, and the tradeoffs they may present in the future. I genuinely think it's a valuable research direction to start thinking about now, especially w.r.t. harness design. alexzhang13.github.io/blog/2…
51
257
26
2,391
289,772
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
4,035
8,297
6,819
76,642
40,094,401
Carson Kahn retweeted
first open weight thinking machine model!! 975B total, 41B active trained on 45T tokens, 1M context, multimodal in sliding window with a 5:1 ratio and 512 size, deepseek aux-free load balancing and 2 shared experts (usually people only use 1), actually curious why the model is less sparse than kimi (~4.2% vs 3.2%). they use a short convolution after k and v, output and ffn (see plot), muon (they cite manifold muon but mention weight decay so not sure), muP, and have a very nice RL scaling curve and chain of thought! one very cool part of the release imo is how well their small variant (276B total, 12B active) performs compared to the big one. they mention they changed the pre-training data mix and recipe, very curious about those changes and to see them scaled up to ~1T (or more?) soon 👀 > "It is not the most performant model available today, closed or open. We trained Inkling for solid capabilities across the board rather than state-of-the-art performance in a single area, to serve as a foundation for the models we will train in the future." also this is really refreshing to see in a model release, huge congrats :)
Today, we are introducing Inkling. Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights available. thinkingmachines.ai/news/int… Available today for fine-tuning on Tinker. Play with it in the Inkling Playground. 🧵
20
48
5
547
150,916
Carson Kahn retweeted
Salesforce
@salesforce7 Nov 2025
Salesforce has signed a definitive agreement to acquire @SpindleAI, a leading agentic analytics platform. The acquisition brings aboard deep expertise in multi-agent systems and high-performance data applications to accelerate Agentforce: sforce.co/4hLOpRA
3
18
5
46
10,624
Carson Kahn retweeted
Learn to ship. Shipping is a skill distinct from coding. Shipping is designing, coding, QAing, story-telling, teaching, marketing, selling, pivoting, iterating… It used to be that coding dominated in importance because of coding ability scarcity. AI will push you to go further.
239
962
193
8,073
1,044,008
Happy to announce that spindle.ai has accepted an invitation to become a @Microsoft AI Cloud Partner! As former @GoldmanSachs CTO and enterprise #AI leader Dr. Paul Walker notes, “Spindle has uniquely delivered agentic AI for our hardest planning and business analytics problems, and I’m thrilled to see this technology become more accessible to Microsoft’s ecosystem of customers.” With contributors from both @MicrosoftAI and @SpindleAI having been at the forefront of #agents since their respective early work on the #GPT family of LLMs, we’re now opening the door to bring rigorous, agentic business analytics into the platforms enterprises already rely on — @Microsoft365 & @Azure. “We’re proud that the trust of enterprises embodied in our stage-leading SOC 2 Type II attestation can now extend globally through Microsoft’s class-leading AI Cloud Partner program,” adds Spindle co-founder & AI leader @CarsonKahn, “giving customers confidence that agentic business analytics is secure, transformative, and enterprise-ready.” 👉 More big things ahead — find our teammates throughout @Workday’s #WDAYRising conference this week (lnkd.in/ghHUfDBD) for a preview of what’s to come. If you’re attending, reply here so we can connect! – at San Francisco, CA
1
2
274
Carson Kahn retweeted
wild what a few matmuls can do
150
112
37
2,092
314,158
Carson Kahn retweeted
Grateful to be alive, for my friends, my partners, my family, and for the opportunity to pursue what I love daily.
4
3
100
6,132
The thing about o1 is the 1 part. Not 2, 3, 4, but 1. We’re still early.
9
7
1
162
16,255
I love asking people what they have unreasonably high standards for. The answers are always very revealing. And if they don’t have an answer—well, that’s also revealing. So, what do you have “unreasonably”high standards for?
488
110
268
1,796
863,995
the first time i read, ``` if __name__ == '__main__' ``` i almost decided to return the book on Python to B&N
28
35
4
1,166
91,554
Carson Kahn retweeted
Selling software will evolve to selling agents, AI that acts on behalf of users. The efficiencies for rote work are too massive to ignore for many uses. As the technology rapidly evolves, so too will the sales strategies. Sellers & the startups they represent will need to re-imagine roles. In a sense, selling AI agents is analogous to category creation. About a decade ago, Nick Mehta & the Gainsight team created the customer success category. In 2014, I remember joining a panel at the Gainsight annual conference Pulse, then in its second year. The conference didn’t highlight Gainsight’s software. Instead, the speakers & the content  equipped customer success leaders with the statistics & stories to convince their peers & managers that customer success needed to exist at their companies. Once budget had been allocated, Gainsight became the de facto software to implement the discipline. Gainsight’s category creation set the standard for many future categories ; Monte Carlo in data observability is another. Just as these companies dreamt new ways of working & evangelized the future, AI agent startups must do the same for their buyers. Working with AI will metamorphose day-to-day workflows, just as much as a new software category. But technology & efficiency aren’t robust enough propositions alone to close a transaction. The fear of losing a job to AI is enough to stonewall its purchase. “AI is Coming for Your Job” is a pervasive & petrifying meme. Buyers facing a threat to their livelihood will block the sale - a halting act of self-preservation. To mitigate this, sellers must help the buyer envision a future with less rote work, greater impact on an organization, & a radically different path-to-promotion than any pitch in the last twenty years. The companies who master this skill will be the early leaders of the next era of SaaS.
15
23
3
178
31,339
Carson Kahn retweeted
Wild to think that a year ago today, ChatGPT hadn’t come out yet, and the few who were playing w openai models were just finding out that davinci-003 could rhyme
42
104
14
1,136
110,790
Carson Kahn retweeted
Me and TypeScript every single time xitter.cf/ciksgibiyim/status/171…
58
336
72
3,347
487,768
Carson Kahn retweeted
Andrej Karpathy
@karpathy22 Sep 2023
LLM knowledge is a lot more "patchy" than you'd expect. I still don't have great intuition for it. They learn any thing in the specific "direction" of the context window of that occurrence and may not generalize when asked in other directions. It's a weird partial generalization. The "reversal curse" (cool name) is imo a special case of this.
Does a language model trained on “A is B” generalize to “B is A”? E.g. When trained only on “George Washington was the first US president”, can models automatically answer “Who was the first US president?” Our new paper shows they cannot!
155
317
37
2,927
876,165
Carson Kahn retweeted
Long-term investments are extremely painful to make, but extremely worthwhile when they finally come to fruition.
40
137
10
1,645
232,412
Carson Kahn retweeted
It's finally out! Been working on a blogpost with a couple LP friends on what good/great TVPI and DPI look like before fund term Simply put, how much of selling positions as a VC is a science and how much is art 👇
7
20
9
153
63,220