Karthik Ramadoss @KIntheHouse

Anthropic starts localizing Claude pricing for India, its biggest market after the US techcrunch.com/2026/07/13/ant…

Super helpful tips. If you have Codex or Gemini subscriptions, you can also have them call those for steps that would benefit from model diversity.

ClaudeDevs @ClaudeDevs

2 weeks ago

A few patterns we frequently use with Fable 5: Use Fable 5 as an "advisor." An executor (Sonnet 5) calls Fable 5 for guidance. Most tokens are billed at the lower executor rate.

728 2K 31K 5.8M 28K

What in the world is forward deployed engineering? The demo went great. Everyone clapped and the pilot got funded. 6 months later it’s still a pilot, and nothing about the business has changed. In the span of 3 days, the 2 largest cloud vendors on earth said the same thing with 9 zeros behind it. open.substack.com/pub/aiwanderlu…

Yay!

Claude @claudeai

2 weeks ago

We're extending access to Claude Fable 5 on all paid plans through July 12.

6K 9K 82K 22.6M 6K

We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5. We'll begin restoring access tomorrow, and will share an update soon. We’re grateful to our users for their patience, and to everyone who worked with us on redeploying the models.

Introducing Claude Sonnet 5, our most agentic Sonnet yet. It makes plans, uses tools like browsers and terminals, and runs autonomously at a level that just a few months ago required larger and more expensive models.

In the first article I traced how every agent capability now gets copied within days and donated to a neutral standard within a year, and how each time that happens the moat doesn’t die, it slides down a layer: from features, to the trust boundary, to identity. I made one claim there that sounded like hyperbole until the dates sit next to each other: the gap between one lab shipping an agent capability and a rival shipping the copy had collapsed from about a year in 2023 to about a day in 2025. The one-day example was payments. Two card networks shipped competing agent-payment standards twenty-four hours apart, and I used it as the punchline. This article takes that punchline seriously and examines the payments timeline step by step. Evolution of Agentic AI: The Payments Standards War Is Already Over (Part 2) open.substack.com/pub/aiwanderlu…

As an Enterprise leader it’s very hard to keep up with the AI related news. To see the signals from noise. I write a weekly distillation of what’s relevant. What I’d read. Every Monday. Share your thoughts if you find this useful. substack.com/@aiwanderlust/…

Check out the new weekly! Oh Mythos is out.. not for everyone though! Read on.. open.substack.com/pub/aiwanderlu…

In January 2025, OpenAI shipped Operator — an agent that drives a computer the way a person does. Three days later, Alibaba's Qwen team open-sourced the same trick. Both landed ~90 days after Anthropic shipped "computer use." Same capability, three names, one quarter. So I pulled the first-party ship date for every agentic capability the major labs have shipped since 2023: function calling, memory, computer use, deep research, connectors, payments. The pattern repeats every time: → One lab ships it first. → Rivals ship a relabeled copy within months, sometimes days. → Then it either gets donated to a neutral standards body and stops being anyone's advantage — or it quietly dies. The dying part is easy to miss. Plenty of celebrated "firsts" are now footnotes. The copying isn't the interesting part. The clock is. In 2023 it took Anthropic 352 days to match a capability OpenAI shipped. By 2025 that gap was down to 53 days. For agent payments, two card networks shipped competing standards one day apart. And no, MCP didn't cause it. The fastest copy in the whole dataset finished four months before any rival had even adopted MCP. The real engine: open weights you can distill, published recipes, and talent that walks between labs. If you hold a budget, the 90-day clock is a vendor's clock, not yours. It governs how long the company selling you a feature gets to charge a premium for it. Google was a full year behind on computer use and lost nothing — which tells you a buyer can almost always afford to be late too. So stop paying a premium for "first." The thing worth buying is the part that doesn't commoditize: who owns the trust boundary your agents run inside, and who can prove an agent acted for a specific person within limits. This is Part 1 of a three-part series. One question for your next architecture review: list the agent capabilities your vendors are selling you on, then assume every one is free and standardized by next quarter. What are you actually paying for? If you can't answer that without naming the capability itself, you're paying a premium for "first." What's the first agent feature your team paid up for that's now table stakes? Curious where this lands for you. Evolution of Agentic AI: What It Tells You About the Next 6-12 Months (Part 1) open.substack.com/pub/aiwanderlu…

x.com/i/article/2070…

3rd gen AI UI/UX! I like that description @karpathy

Andrej Karpathy @karpathy

a month ago

This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide. Once you do all of the under the hood engineering work to make this "just work" (e.g. across tools, integrations, compute environments, memory,

1K 2K 23K 8.0M 14K

AI Waypoints #15 is up. The AI buildout got physical this week. Amazon confirmed it's in talks to sell its Trainium chips outside AWS — a direct shot at NVIDIA. FERC ordered every US grid operator to fast-lane power for AI data centers. And Accenture's earnings showed AI eating the consulting work it sells. 🧵 AI Waypoints: Week of June 22, 2026 — Edition #15 open.substack.com/pub/aiwanderlu…

Over the last two weeks, both the U.S. Government and Anthropic took significant actions that demonstrated their power to control access to AI by restricting what others can do with frontier models. This has been one of those moments that, once seen, will be hard to unsee, and it is significantly accelerating many businesses’ and nation states’ efforts to ensure reliable access to AI that no one else can terminate. Anthropic first released Claude Fable 5, a version of its Mythos model with additional guardrails, including some restrictions that seem well justified on safety grounds (such as limitations on applying it to hacking, bioweapons, and so forth). However, it also restricted developers’ ability to use it to build competing LLM technology. This move was concerning, given that the whole AI community, including Anthropic, has benefitted tremendously from open research — indeed, the AI revolution was kicked off by my former team (Google Brain) freely publishing the Transformers paper! Imagine if Microsoft’s terms of use barred anyone from using their tools to build competitive software, or if Google barred using it to search for information to work on competing search engines. Anthropic’s argument that it was unsafe for others to be able to make advances in AI also rang hollow. Initially, Anthropic silently degraded Fable 5’s performance for users detected to be working on LLM research through invisible interventions that weakened the model’s outputs without notifying the user. After significant backlash, it walked back this decision and decided to be transparent when it did this, but it still refuses to use its latest capabilities to help AI researchers. This move represents a raw demonstration of power by Anthropic. It has used “safety” arguments to hinder potential competitors. Platforms succeed when they are viewed as stable, reliable partners that one can build on. The sudden rule changes by Anthropic (including a mandatory 30 day data retention policy for Fable usage) have made developers wonder about the stability of building on any one proprietary LLM provider, not just Anthropic. The U.S. Government then shortly followed with an even greater demonstration of power. It used the Commerce Department’s authority to regulate technologies that may be national security threats to restrict exports of Mythos and Fable, requiring a license for use by any foreign national, whether inside or outside of the U.S., including employees of Anthropic. This led Anthropic to disable access to Fable to all users worldwide. Sam Altman pointed out, referring to Anthropic, “It is clearly incredible marketing to say, ‘We have built a bomb, we are about to drop it on your head. We will sell you a bomb shelter for $100 million.’” But when one engages in this type of fear-based marketing, it increases the odds that the U.S. Government will agree with you and slap export controls on the bomb you say you have built. To be clear, I don't think Anthropic has built anything like a bomb, and I don't think export controls on Fable are appropriate. However, following the U.S. Government making this move, many nations, including U.S. allies, saw how the U.S. can suddenly yank their access to AI models. In many capitals around the world, this has spurred discussions on AI sovereignty and how others can ensure uninterrupted access to this critical technology. For decades, many nations were comfortable having many parts of their supply chain rely on the U.S., China, and other major producers. Once a nation issues a threat, or takes action, to limit other nations’ access, other nations will rationally try to secure alternatives. For decades, semiconductor manufacturing in China made slow progress; once the U.S. moved to limit China’s access, China’s efforts kicked into high gear. Similarly, once China threatened U.S. access to rare earth minerals, U.S. efforts to secure alternatives accelerated. Now that it has become crystal clear that private U.S. companies and the U.S. government can limit, in short order, other nations’ access to frontier AI models, the incentive of others to invest more in alternatives like open source grows significantly. Of course, training frontier models is not easy, so it remains to be seen how successful they are, but we have crossed the rubicon. Satya Nadella wrote an essay about the importance of building a healthy ecosystem on top of frontier AI technology. I heartily agree with him, and hope this week’s events will ultimately prove to be constructive steps toward this. I hope we can build a more free, more open world, where research is freely shared, and laws and societal norms shape a level playing field that allows everyone to make progress. A silver lining of the events of these past two weeks is now that everyone better realizes key points of instability of the current system, we can all work to create a more stable foundation. [Original text: The Batch newsletter]

Football spent 15 years making the calls more accurate than ever: a sensor in the ball, cameras on every limb, 3D replays, refs announcing calls over the loudspeaker. Fans have never been angrier — 8 in 10 want it gone. Accuracy up, trust down. The referees are finally right. Everyone hates it. open.substack.com/pub/aiwanderlu…

Monotheist vs Polytheist Every leader has made this hire. You can bring on one brilliant generalist who holds the whole problem in their head, or a team of specialists with someone to coordinate them. You already know how each one fails. The generalist becomes a bottleneck the day they’re out sick. The team spends Tuesday in a standup about Monday’s handoff. This spring, in Agentic AI, that same hire became an architecture decision, and the two biggest names in enterprise software made it in opposite directions. I’m calling the camps monotheist and polytheist because it stuck in my head and it beats “single-agent versus multi-agent.” One god-model that does everything, or a pantheon of smaller gods, each with a domain, coordinated by a planner above them. Monotheist: one generalist agent, one continuous context, working the task on a single thread. If you need to handle more work, you run more copies of that one agent. OpenAI’s Workspace Agents is the picture: one thread, dozens of apps, no orchestrator. Polytheist: an orchestrator that decomposes the request and routes the pieces to specialist agents, then assembles the result. SAP’s Autonomous Enterprise is the picture: fifty coordinators, two hundred specialists, no generalist. Those are the two choices. The rest of the decision is about which one fits the work in front of you. Monotheist vs Polytheist: How to Choose Your Agentic AI Architecture open.substack.com/pub/aiwanderlu…

The State of Brain-Computer Interfaces in 2026 Three things happened in the past twelve months that change the BCI picture for 2026. Apple released a native protocol for BCIs in iOS. Meta wrapped a six-year, billion-dollar moonshot by launching a wristband that reads muscle signals on the forearm. Vermont’s governor signed a bill recognizing “freedom of thought” as a legal category. Those are three different stories moving at three different speeds, and the news cycle keeps mashing them together. The question for 2026 is which one lands on your desk. Disclaimer: I’m not a neuroscientist. I went looking for an answer to one question: what should an enterprise leader pay attention to when they hear “brain-computer interface” and the first instinct is to file it under hype? The State of Brain-Computer Interfaces in 2026 open.substack.com/pub/aiwanderlu…

This is a super exciting release - Claude Fable 5 is the same underlying model as Mythos but with added safeguards. The benchmarks are great and it's SOTA on everything by a margin but I'll add that *qualitatively* also, this is a major-version-bump-deserving step change forward (imo of the same order as Claude 4.5 was in November), peaking especially for long problem-solving sessions on very difficult problems. You can give it a lot more ambitious tasks than what you're used to, the model "gets it" and it will just go, and it's never felt this tempting to stop looking at the code at all (but don't do this in prod!). The model still has quirks that people will run into and the safeguards are configured to be a little too trigger happy for launch, which can hopefully be tuned over time. I feel a lot of things changing as working software increasingly comes out on a tap. The Jevon's paradox kicks in and I feel my own demand for software growing substantially. You can ask for anything - explainers, visualizers, dashboards, bespoke single-use apps (e.g. a full wandb that is hyper-specific just for your project), you can 10X your test suite, auto-optimize code, run giant research projects with custom HTML for the results, anything! "Free your mind" (Matrix ref). Really looking forward to all the things people build!

Claude @claudeai

a month ago

Fable 5 is state-of-the-art on nearly all tested benchmarks, with exceptional performance in software engineering, knowledge work, scientific research, and vision. The longer and more complex the task, the larger Fable 5’s lead over our other models.

510 2K 16K 5.9M 2K

An Enterprise Leaders monday morning catchup on AI news - no noise, pure signals. The week the AI supplier map got redrawn twice — Microsoft launched 4 of its own top-tier models, OpenAI landed on AWS, Anthropic filed to go public AI Waypoints: Week of June 8, 2026 — Edition #13 open.substack.com/pub/aiwanderlu…

AI Waypoints #11 is up. This was the week enterprise AI's bills became line items — the hyperscaler invoice, the consulting invoice, and the severance schedule. Three of seven signals touch Anthropic. AI Waypoints: Week of May 26, 2026 — Edition #11 open.substack.com/pub/aiwanderlu…