Ad Astra @elonmusk and @SpaceX 🫡🚀

Most of @sequoia went on a hardware trek to LA this week We visited @reflectorbital @Nominal_io @Mach_Industries @neros_tech (we got a drone show; last photo) @K2SpaceCo @VardaSpace @GoToImpulse (And more) The ecosystem is buzzing HARDWARE IS BACK, baby 🚀

Did a podcast w @dan_biderman, @realJessyLin and @sonyatweetybird The @EngramLab team is cracked!

Today's AI models train once. We don't work that way. We learn continuously, forget what doesn't matter, and retain what does. That gap is what @dan_biderman and @realJessyLin are closing at @EngramLab. AI that never stops learning, with memory that lives inside the model instead of bolted on as an afterthought. In our latest Training Data episode we get into why memory is the next frontier: why the brain forgets on purpose, why RAG is a band-aid, and what becomes possible when a model is always training. 00:00 Introduction 00:59 Always Training Explained 01:51 Beyond Context Windows 03:29 Ngram Product Overview 04:34 Adapters And Training Signals 05:32 Internalize Vs Externalize 06:49 Compute And Token Savings 08:19 Teams First Then Individuals 08:51 Memorization Vs Understanding 12:47 Dreams And Offline Digestion 14:08 Training Beats Curation 15:19 Why Everyone Needs A Model 21:44 Bitter Lesson And Architecture 24:44 RAG Killer And KV Cache 31:38 Future Of Memory And Models

“Tech is largely a [...] tinkerer culture. It's not even science, it's tinkering. In science at least you respect the history. If we ask people who Douglas Engelbart or Alan Kay the computing pioneers are, most people in tech have no idea. So tech is like an industry that doesn't know its past. If you don't know your past, you don't know history, which is humanity. Your point of view is like what's in front of you, what is your competitor doing. Versus if you bring in humanity, you have all other disciplines around you. You have all the history behind you. There's way more good stuff you can steal.” - @ivanhzhao

Congrats @neilmovva, @blintzbase, and the whole @sailresearchco team! Tokens have fast become the lifeblood of every AI startup. @sailresearchco is building the infrastructure to scale agents to the stars and back. Let the tokens flow!!

Introducing our first model, Un-0! We trained an image generator powered by a backbone of coupled oscillators in place of a more traditional conventional neural network.

The GTM stack has been overdue for a reckoning. @mutinycorp just replaced it with a single agent. It takes accounts from cold to closed. @jalehr and team have been building toward this for months — this is the real deal.

I built 𝗛𝗮𝗹 - my AI agent/second brain trained on my entire life. He now runs my day, sends emails, organizes projects, collaborates with the team without me, and 𝗻𝗼𝘄 𝗽𝗮𝗿𝘁𝗶𝗰𝗶𝗽𝗮𝘁𝗲𝘀 𝗶𝗻 𝗺𝘆 𝗭𝗼𝗼𝗺 𝗺𝗲𝗲𝘁𝗶𝗻𝗴𝘀. He’s not just a copilot anymore. He is me. See video 👇 This is V1 - Hal is starting to move to autopilot. Next: he’ll run more of my meetings and work without me. He already remembers more than I do and he’s getting smarter fast. At some point, I imagine him handling more of my meetings and work, without me. He remembers more than I do and will be smarter. What capability should I build into Hal next? (He’ll be animated by next week — video 👇)

Thrilled for my friend @dan_biderman to launch @EngramLab Dan is a visionary who has consistently let seen around corners in AI

At @sequoia we help founders building legendary companies. My favorite kind of founder has deep, authentic connections to the industry they're transforming. @georgeprobook is that kind of founder: George grew up the son of a NYC cop. His dad got injured, forcing retirement. He got into the trades and started power washing houses, bringing George with him. George spent much of high school in the truck, learning the trade. He was a smart kid. He earned a perfect 36 on his ACT and got into the Ivy League, studying the prestigious M&T program at University of Pennsylvania. But he did not desert his roots. He kept working with his dad through college, and worked at a home services shop in Illinois that later became Probook's first customer. Eventually, he took some of his AI skills he learned at college to automate the kind of work his dad hated doing: scheduling, logistics, back-office. Then, he offered it to other operators like his dad. Before he knew it, Probook became a smash hit in the home services industry. George had technical chops- but he needed someone who could take Probook to the next level. That's @probookcto . Lewis is a genius. This is a guy who earned a perfect score on the American Mathematics Competition. That isn't a top 1% thing, that is something only 10 out of 45,000 top math competitor achieve. Then, he went to UC Berkeley where he proceeded to early an Electrical Engineering Computer Science (EECS) degree with a 4.0 GPA...in 2 years flat! I heard phenomenal reviews about Lewis from my friends at Roblox, where he was top of the pack for young engineers. But most remarkably, Lewis is truly a ray of sunshine. Go look at his web profile. You'll see him smiling. He ALWAYS has that smile. Like 100% of the time. Even during the hardest moments. He has such a strong mental fortitude (and back strength...man he can do pullips) that he really inspires me. Probook's third and final co-founder, Ben Cervantez, grew up in SF where he was Valedictorian, Captain of Varsity Football, played Varsity Baseball, and earned a perfect 36 ACT. Yes, the full package. He Attended Wharton's prestigious M&T Program where he played Men's Rugby. But when he learned about George and Probook, he took a bold move and dropped out to go full-time. This was extremely high risk, but clasically Ben's style. He's low ego and high performance. He always puts the team first. Ben Cervantez is the glue that holds this team together. He works tirelessly, flies non-stop to customers, and lives with extreme customer love. Because of the authentic connection and customer love, customers love Probook. As one customer wrote, "If you take Probook away. I quit." The growing like crazy, hiring an incredible in-person team in NYC, and really serving the trades industry with care. If you can't tell, I really love working with this team. I love being their "coach." These daring founders are the real deal in every way. Congratulations to my friends and partners on the fundraise!!

- 26,000 agents affected - gained full control of every one of these agents, their private conversations, and every internal system the agent could reach. - passed all existing security scanners - got 37,000 real GitHub stars ^^ Those are the shocking results of an experiment that @YairS17728, @niv_hoffman & team ran. air.security/blog-posts/the-…

The advanced civilizations of sci-fi legend (Banks, Asimov, etc) have some form of simulation to guide society. @joon_s_pk is taking a crack at building that simulator with @simile_ai. As Joon's cofounder @percyliang puts it: great science starts with a great measurement. From Smallville in 2023 to today, Simile is trying to build the Hubble Telescope equivalent for simulating human behavior. 00:00 Introduction 01:49 Building Generative Agents 02:29 Valentines Day Emergence 03:33 From GPT 3 To Agents 05:03 Social Computing Problem 06:19 Social Simulacra Subreddits 07:57 Models Getting Good Enough 08:57 Humans Are Not Rational 10:04 Turning Research Into Simuli 11:55 Validation And Accuracy Proof 12:43 Customer Workflow CVS Example 16:11 Why Collect Real Data 17:51 Behavioral Signals And RCTs 21:52 Use Cases And Second Order Effects 26:31 Evaluating Convergence Divergence 31:58 Big Societal Simulations Ahead 36:08 Future Of Simulation

"When you start a company, the expectation is failure." You can't build a great product. You can't hire great people. You'll never get good investors. You can't raise a Series A. You can't raise a Series B. @mutinycorp CEO Jaleh Rezaei (@jalehr) had heard all of those doubts. She'd done all of it. And she started over anyway. "It's really hard to accept that you're going to burn that to the ground for the possibility of a better future," she says. After a year of working to add new AI capabilities to Mutiny's software, Jaleh felt she was "mashing" two different products together. On a walk along the Hudson River in November, she told co-founder Nikhil Matthew they had to make a radical change. Catch the full story on The Upstarts Podcast, presented by @Rippling 🫡

Salespeople have a problem: they only spend 21% of their time actually selling to customers 🤝 @mutinycorp CEO Jaleh Rezaei (@jalehr) thought she'd found a way to fix it. She worked with customers like Snowflake and Uber. Revenue reached 8-figures 📈 Which makes what happened next all the more shocking: she burned it all down 🔥 On The Upstarts Podcast, Jaleh shares why she took the tough decision to fire her customers, slash headcount and re-found her startup around an AI agent. We go into AI's disruption of SaaS; her favorite advice from @fin's @destraynor on pivoting and from @ycombinator's @garrytan on showing the chart you don't want people to see; how to get over "Claude Spookies," and why speed is (mostly) all you need. This season of the show is presented by @Rippling 🫡 CHAPTERS 00:00 Introduction 2:41 An AI agent for sales 7:04 Why an AI ‘blank canvas’ doesn’t work 10:38 Jaleh’s founder journey and Gusto lessons 17:01 Mutiny’s early software success 22:51 Jaleh’s Upstart Moment: tearing it down 24:42 Why you can’t rebuild a business halfway 27:15 Intercom’s founders share valuable advice 30:44 Focus on the business, not the optics 34:53 ‘Speed is the only thing that matters’ 37:40 ‘Claude spookies’ in the AI app layer 43:59 Second-time founder advantage Thanks for watching!

12 years ago a top scientist said AI would wipe out radiology. Computer vision spread through every scan – and radiology demand went UP. Hospitals hired more radiologists. @Nvidia’s Jensen Huang on why we keep confusing a job's tasks with its purpose.

Don't mistake task for purpose. @Nvidia's Jensen Huang in conversation with @Konstantine: "Typing is not the job of a software engineer. Coding is not their job. Solving problems is their job." Jensen says AI is creating more jobs, not destroying them. Why? AI lets us solve problems better than ever.

Last week Apple previewed the future of Siri. In 1987 though, Apple showcased a far more advanced AI assistant that would change how we use our computers entirely. It could see you, control your computer, and even looked and sounded human. They called it Knowledge Navigator. For nearly 40 years it remained science fiction. This past week at @tavus we finally brought it to life with the help from our friends at @cerebras. Meet Dom:

Sequoia partner and SpaceX investor Shaun Maguire tells Ed Ludlow why he "stands incredibly bullish" on his orbital data center outlook for SpaceX bloom.bg/4ouhhRI