| |
Welcome, humans. | Yesterday on TBPN, Salesforce CEO Marc Beinioff called off the SaaSpocalypse, called out the fact that AI lab Anthropic uses seat-based pricing in its enterprise plan as the ultimate bull case for the longevity of Salesforce's own seat-based model (such a good point), and also made some hilarious dolphin noises, and while he was at it, some whale ones too. | Meanwhile, Martin Peers over at The Information broke down the good news, bad news of the Salesforce earnings report. So...SaaSpocalypse just on hold, I guess? | Speaking of putting things on ice (or should we say putting ice on things!)… we've got some good AI news for folks who've been dealing with blizzards lately: robots who can shovel now! | | Is this the most efficient form factor? Absolutely not (maybe you'd be better off with the equivalent of a roomba with ski-wheels and a snow-plow on the front of it). But is this very entertaining? Absolutely yes. | Here's what happened in AI today: | NVIDIA reported record Q4 revenue of $68.1B (up 73% YoY) and guided $78B. Anthropic dropped its Responsible Scaling Policy for more flexible terms. Perplexity launched Perplexity Computer, a system that orchestrates 19 specialized AI models to handle complex tasks simultaneously in the cloud. Someone jailbroke Claude to steal 150GB of sensitive Mexican government data.
| | LATER TODAY at 10AM PT / 1PM ET: We're going LIVE to talk to Daniel Reid Cahn, whose company Slingshot AI built Ash, an AI purpose-built for therapeutic support from the ground up. |  | Click the image above, then on YouTube, click "Notify Me" to get notified when we begin. |
| Millions of people are already using chatbots for emotional support every day, whether the companies behind them designed for it or not. Daniel thinks there's a better way. If you're using AI for emotional support (we know a lot of you are), don't miss this one. | | Perplexity wants to replace your computer with 19 AIs, and we're… kind of here for it?? | FULL BRIEF: Perplexity Computer: 19 Models, One System, Everything to Know | You know that thing where you start a project in ChatGPT, realize you need an image so you jump to another tool, then need code so you open something else, then need research so you're back to square one? It's like cooking dinner using a different kitchen for every ingredient. | Perplexity just launched something that's trying to fix that. It's called Perplexity Computer (try it here) and the pitch is simple: instead of one AI model doing everything okay, 19 specialized models work together, each handling what it does best. | Here's how it works: you describe what you want done (say, "build me a competitor analysis dashboard"), and the system breaks your request into subtasks. A reasoning model plans the approach. A coding model writes the code. A research model pulls live data. An image model handles visuals. All running simultaneously, in the cloud, while you do other things. You can watch a series of examples here; they're wild! | CEO Aravind Srinivas quoted Steve Jobs to frame the vision: "Musicians play their instruments. I play the orchestra." Well, now YOU play the orchestra. Get your conductor wand (stick? baton? giant rubber hammer for agent Whac-A-Mole?) ready! | Early demos are impressive: | | Here's what's actually under the hood: | Claude Opus 4.6 handles core reasoning and orchestration. GPT-5.2 manages long-context queries and web search. Grok runs lightweight, fast tasks. Veo 3.1 processes video; Nano Banana generates images. Gemini powers deep research.
| The system also has persistent memory, meaning it remembers your past projects, files, and preferences. Tasks can run for hours, days, even months, without you babysitting them. And if you're reading this thinking, "Gee, this sounds a lot like OpenClaw"; you're right, it's a lot like OpenClaw… but a whole heck of a lot easier to set up! | It's available now for Max subscribers ($200 / month), with Pro and Enterprise access coming soon. Max users get 10,000 monthly credits plus a 20,000-credit launch bonus. | The bigger picture: this is Perplexity's bet that the future of AI isn't one model that does everything. It's a team of specialists, orchestrated automatically. So now you can stop asking "which AI / tool should I use?" and focus on "what task do I want to get done?" This is a recurring theme recently: Agents getting ACTUALLY useful. | |
|
|
|
|
What the latest DORA report actually says about confidence in AI Why the "productivity paradox" is creating friction between engineers and the org How teams replace "trust" with feedback loops
| Register with your questions, and leave with a renewed sense of "AI-trust". | Grab Your Spot! | |
Prompt Tip of the Day | You can now tell Claude to do something once and have it run automatically—every day, every week, whenever. Scheduled Tasks in Cowork lets you type /schedule, describe a task (e.g. "summarize yesterday's Slack messages"), pick a cadence, and Claude handles it on autopilot with full access to your Google Drive, Slack, and plugins. Watch the walkthrough. This is where your prompts get replaced by automations. | Meanwhile, for engineers, Claude Code Remote Control lets you run claude remote-control in your terminal, then continue your session from your phone, tablet, or any browser. Claude keeps running locally, and you can steer it from anywhere. | There's one catch w/ scheduled tasks: they need your computer awake + Claude Desktop open. Also, Remote Control needs Pro or Max. | | Treats to Try | *Asterisk = from our partners (only the first one!). Advertise to 650K readers here! | |
|
| Around the Horn |  | This is probably the most important ongoing story in AI right now… Not that we find this idea at all funny, but because of course, someone DID make a pretty banger meme of "WarClaude" |
| Anthropic dropped its 2023 Responsible Scaling Policy requiring pauses when AI capabilities outpaced safety measures, replacing mandatory "red lines" with flexible responses that only delay development if Anthropic leads the race and risks are catastrophic. Amazon will reportedly invest up to $50B in OpenAI, with $35B of that contingent on OpenAI either going public or reaching AGI. NVIDIA reported record Q4 revenue of $68.1B, up 73% year-over-year, driven by $62.3B in data center sales, and guided for $78B next quarter while unveiling its next-gen Rubin chip platform (video; very cool) promising 10x cheaper inference than Blackwell. A bad actor jailbroke Anthropic's Claude to identify vulnerabilities, write exploit scripts, and steal 150GB of sensitive Mexican government data including taxpayer records and voter information over a month, prompting Anthropic to ban accounts and enhance anti-misuse tools. DeepSeek withheld its upcoming V4 model from U.S. chipmakers like Nvidia and AMD, giving early access to domestic suppliers like Huawei instead amid U.S. export controls and allegations it trained on Nvidia's Blackwell chips in China. Companies including Amazon (16K cuts), WiseTech (2K), HP (4-6K), Dow (4.5K), and Allianz (1.8K) announced thousands of layoffs as investments shift to AI, with Goldman Sachs estimating AI drove 5-10K monthly net U.S. job losses in exposed sectors. A Chinese official's use of ChatGPT as a personal diary accidentally exposed a global intimidation operation targeting dissidents abroad via impersonated U.S. officials, forged court documents, and thousands of fake accounts; OpenAI banned the account. Google DeepMind's Aletheia solved 6 out of 10 FirstProof math problems autonomously using Gemini DeepThink, marking superhuman-level reasoning progress on advanced mathematics. Google is likely to publish Nano Banana 2 tomorrow; how do we know? Logan Kilpatrick tweeted a banana. Logan Paul made a 15 minute AI movie with the DOR Brothers, aiming for a production value of $300M but made in only 7 days.
| P.S: We had to skip Thursday Trivia today. We'll do it tomorrow, for Friday trivia! | To read EVERYTHING that happened in AI this week, click here. | |
|
|
|
Test-drive the self-guided demo and see how quickly you can create. | |
|
| | | That's all for now. | | What'd you think of today's email? | |
|
| P.P.S: Love the newsletter, but only want to get it once per week? Don't unsubscribe—update your preferences here. |
|
No comments:
Post a Comment
Keep a civil tongue.