Today, I’m proud to announce Sardine AI Labs, an applied research group building fraud and financial crime AI models. As part of the launch, we’re funding 5 independent researchers with $375,000 in total funding. I’ve wanted to build this for a long time. Let me give you an early example of the work we’re doing: Typically, new card programs lack the data to detect fraud reliably, and they're often among the first targets for fraud rings. In testing on card issuers held out of training entirely, the very first foundation model from Sardine AI Labs improved fraud detection accuracy by 68% for a consumer card issuer and 41% for a business card issuer compared with standard machine learning approaches. Our consortium recently crossed 6.5B devices profiled and $1.8T in payments. This data, combined with our fantastic team of AI researchers and engineers, makes this lab and these early results possible. In addition, the Sardine AI Fellowship Program is an initiative I wanted to launch alongside Sardine AI Labs. When I was writing my PhD dissertation in the early 2000s, the preeminent research labs like Bell Labs, HP Labs, AT&T research did a lot to move research on various internet technologies forward. Over time, the concept of research labs withered away. This is our attempt to give back to the research community. We're funding up to 5 research fellowships, with $375,000 in total, for researchers who want to help push this work further. Key details: - Advanced-degree students, researchers, and faculty at US or Canadian universities. - Fellows work with the Sardine AI Labs team to build new foundation models, and are encouraged to publish in peer-reviewed venues. - Research areas include cross-institution generalization, multi-stream modeling, scaling laws, identity matching, real-time inference, cold-start learning, and applications beyond fraud. Here’s more about Sardine AI Labs and how to apply to the fellowship: sardine.ai/ai-labs I am absolutely thrilled to launch both of these initiatives. Thank you to the entire @sardine team who has brought this vision to life.

Excited to announce our $110M raise to scale our stablecoin-native banking platform for global enterprises across 35 countries (10 new countries!) 🚀🦁 Our stablecoin business crossed $1.5B annualized volume in just 8 months, driven by exponential enterprise demand. Since signing the Series C-1 term sheet, revenue has grown more than 50% in a few months. The pace is accelerating. As we go deeper into the stablecoin ecosystem, I'm pumped to have @coinfund lead the round, joined by @AB_insights, @a16z, @cbventures, @CRV, GIC, @gpt, @paraficapital, Vista, @wmt_ventures, @ycombinator and others. Also grateful to have the best founders building in the space joining the round: @zcabrams (Bridge, OUSD), @psneville (Circle, Catena), @sternhenri (Privy), @prabhakar2reddy & @karan705 (OpenFX), @SamBroner (Better Money Co.) and Ricardo Marino (Global Chairman at Itaú), among others. Over the last year, AI & stablecoins have powered us to an inflection point. Today, we have fewer employees than we did two years ago, while revenue and volume have grown more than 10x crossing $5B in annualized TPV. AI is embedded across Jeeves and has fundamentally changed how we build and operate. Alongside the round, we're launching: 💸 Jeeves stablecoin wallets: Our custom built wallets supporting USDC and USDT, with payouts to 190 countries, all in one platform. 💳 Cards in 10 more countries: Argentina, Peru, Paraguay, Uruguay, Panama, Guatemala and others. The same award winning expense management platform used by BMW, Burger King and H&M are now available in these markets. 🧾 Accounts Receivable (AR): Alongside our core Cards and AP modules, we're launching our AR module, starting with Brazil 🇧🇷 This was one of the biggest customer requests! 📊 Jeeves Global AI Spend Tracker: We all know AI spend is growing but if you operate in six countries, with Claude spend in pesos & OpenAI spend in pounds, how do you actually know what you spent across models, countries, currencies & what drives ROI? Well now you do with our Global AI spend tracker across the globe 🌎 🇪🇸 Lastly, we opened our doors in Madrid to accelerate our expansion across EMEA, with a focus on stablecoin payment corridors between Europe & the world. The next decade of global banking will be agentic and seamless between stablecoin and fiat. To build this companies need to own the primitives directly. This chapter is being written right now, and small teams with big ambition, powered by AI and stablecoins, will write it. We intend to be the company that defines it 🌎 Thank you to our customers, teammates and investors for the trust you put in us every day 🙏 If you want to shape the future of global money at the cutting edge of stablecoins and AI, come build it with us. It won't be easy but it will be fun 👉 tryjeeves.com/us/about-us Day one of this next chapter starts now. It's time to build.

Oh man there is so much in the OAI launches this morning, early reactions: - OAI leaning into its strengths around model breadth (full duplex voice / computer use), compute (ultrafast / decisions / price cuts), and knowledge work (spaces / dots) - Spaces is an impressive effort to connect two key lessons from the last 6 months; one that consumers need "human shaped" interfaces to use models to their full ability (slack / imessage / anthropomorphic agent) and two that group productivity >> sum of individual productivity and requires dedicated collaboration tools like agent native docs / slides / etc. - Dots acknowledges that consumers have a bounded view of chat interfaces ("better google") and the personal agents primitive needs a different positioning and personality. I like that the UI starts with a single "dot" (agent) and imagines expanding to teams, which combines the simplicity of a muse style approach with the power of grok bots, progressively disclosed. - Security cloud is the first of many efforts where sensitive / proprietary model capabilities are vertically integrated and sold as an outcome vs API. Expect more of this in areas like biology (smartly framed today as "antibiotics" - who could dislike those?! 😂). - Yes there are offerings at every spot on the Pareto curve (ultrafast at the high end, decisions at the low), but the wild 3D demo of ultrafast shows that these points on the curve represent new capabilities (i.e. just in time software, real time video transformation etc) - Consumer adoption has accelerated w/ 1.2B WAU + Connect is a huge unlock for developers by letting users bring their own tokens—this unbundles the unit economics from product value & locks developers + users into the OAI ecosystem - Plugin directory + discovery is maybe, finally the distribution channel we've been waiting for - Marketplace allows SAAS partners to sell to OAI customers w/o the procurement headache - $500 plan, and it's "cheap" given the capabilities .. I predict a $1000 plan by next year that will provide even more price/performance to consumers Will think + share more when I have a chance to process, but wow some cool launches today.

We instituted an official AI Writing Policy at @Clay. Massive thanks to @sophiebits on our engineering team who wrote this. Originally this was just for eng, but other teams found it so helpful we expanded it company-wide. Here are the four guiding principles: 1. You must stand behind every idea and sentence It is your responsibility to make sure that the entire document is representative of your own thoughts before you share it. 2. Writing is thinking Spending time on the writing process teaches you more about your topic. If you circumvent this process, you will walk away with a poorer understanding of the subject matter. 3. More time should be spent writing a document than consuming it If you generate a document from a short prompt then ask your readers to go through the longer output, you are disrespecting their time. They can talk to ChatGPT themselves if they want to. 4. Longer is not better AI makes it easy to generate long docs, and it loves padding them with sentences that say nothing. If you're producing docs from a short prompt, consider just sharing the prompt. You can read (and borrow) the full policy below. Hopefully we all communicate a bit more clearly now :)

I don’t normally write these things. But this one hits a bit different. Nearly three years ago I told a16z I wanted to spend part time outside the firm helping @drfeifei start and build a frontier model company. And what followed were some of the most remarkable moments of my career. From being at the founding table. To watching the first git commits. Seeing the very early very rough results that hinted at something much greater. Contributing to the open source. Watching the creation of new model architectures that pushed the state of the art in multiple areas. I saw the team spin magic from nothing again and again. And I was very privileged to be within the walls. I’m very proud of what was accomplished, and am so excited for what lies ahead. So congratulations to AMD for entering into an agreement with the leading spatial intelligence frontier lab. And congratulations to the World Labs team on a partner that has the vision, ambition and leadership to be the top AI technology provider globally. Also congratulations to Dr. Su and Dr. Li. Two world class CEOs with a common vision. The both of you working together is unimaginably legendary. Arguably the smartest leaders in tech with the shared goal of building the world's leading AI capabilities. I know you both share an optimistic view of AIs ability to aid humanity. We need far more of that. As everyone knows. I’m World Lab’s biggest fan. And will remain so. Thanks to @drfeifei , @BenMildenhall , @jcjohnss and the entire team for letting me tag along. What a crazy fucking ride. You stood at the frontier, and moved it. And will continue to. Again, many congratulations to everyone involved. The future of AI will be so much brighter with this partnership. I can’t wait to see where it leads. Here’s to new worlds!

Amazon cuts off Muse. While I am bullish Meta and Muse, I think many people are overlooking the digital knife fight that’s about to occur Nobody wants to get commoditized or layered here. Let the games begin

.@fal's Gorkem Yurtseven and Batuhan Taskaya on making an open source video model 35x faster, and what Hollywood wanted after they built it: Last month, MiniMax released H3, an open source video model. fal rebuilt it - they cut down the steps the model takes to make a video, rewrote the code under each stage, and got the GPUs to 70-80% of their theoretical ceiling instead of their usual 30-40%. No quality loss. Video now generates faster than you can film it. The models have gotten so cheap and fast that end users aren't even asking for improvements in either category anymore. The gap has moved to quality, or how closely the model follows the prompt. Hollywood wasn't a customer a year ago and is now fal's fastest growing segment. Studios love it for the little things - extending a shot, moving the camera, changing the lighting. Those edits land 80-90% of the time. fal is chasing 99.9%. In this conversation with a16z's Jennifer Li: 00:00 Intro 01:45 The first open model worth rebuilding 03:45 The industry ran out of compute in April 06:25 35x faster without losing quality 11:05 5s of video generated in 1.5s 14:25 The weekend fal dropped everything 16:15 3 viral projects nobody planned 18:20 Teaching a video model to remember 20:25 An hour of video that's consistent 23:30 2x the usage of other models in 3 weeks 26:35 The case for giving video away for free 29:00 Blender sketch in, finished shot out 30:40 Chasing 99.9% reliability 34:15 Hollywood, fal's fastest-growing customer 36:50 Why studios wouldn't touch it until now YouTube: piped.video/SDbRJXQrYGY @gorkem @isidentical @fal @JenniferHli

Lazy work used to mean too little output. Now, with AI, it often means too much and more work for everyone else. @tobi call it "slop grenades." A "Slop Grenade" is when you let AI produce the work and pass it on without adding any value (including checking it). Someone else has to wade through it, catch the mistakes, and clean up the mess. You save time and look productive but someone else pays for it.

Yesterday I uploaded my auto insurance policy to Meta @Muse and asked for a better rate with identical coverage. In ~5 minutes, it found a policy saving me $3,500 /year, bought it, and canceled my old one. I knew I was overpaying. I’d abandoned several attempts to switch because insurance websites and aggregator portals are so painful. Why fill out forms, compare quotes, and field sales calls when an agent can finish the entire job in five minutes? Killer product.