Seeing a number of benchmarks showing Opus is the best model for long-running work.

Five tips for running Opus autonomously for hours/days:

1. Use auto mode for permissions, so Claude doesn’t ask for approval
2. Use dynamic workflows, to have Claude orchestrate hundreds/thousands of agents to get a task done
3. Use /goal or /loop, to nudge Claude to keep going until it’s done
4. Use Claude Code in the cloud, so you can close your laptop (easiest way is the desktop or mobile app)
5. Make sure Claude has a way to self-verify its work end to end: Claude in Chrome browser extension for web, iOS/Android sim MCP for mobile, a way to start the full web server or service for backend work

I have had the chance to go through two IPOs (Meta and Dropbox).

Fabulous wealth tends to amplify deeper desires, not create new ones.

The mainstream narrative is: "Early Employees make lots of money and buy mansions and go to the beach".

For some: Yes, they want to chill. That is great. I am happy for them.

But for a lot of folks, it is a chance to do even crazier and wackier stuff. Start new things. Fund new things. Keep the crazy loop of Silicon Valley going.

It's going to be awesome to have a lot of folks with liquidity in the coming months.

Onwards!.

Vercel AI Gateway recovers on average over 1T tokens a month 🤯

Much like Stripe recovers revenue with smart retries on failed payments or credit card updates.

And we do it with 0️⃣ zero markup over the labs; adding redundancy, zero-data retention enforcement, observability, usage APIs, caps, …

https://t.co/OougSipbBX

Replit is about removing all distractions and have you focus on what matters — getting to market and getting the bag. Congrats! https://t.co/Z2gcG0WF0t

The numbers may be a bit extreme here, but unquestionably use-cases have to stratify in the next year or two between model families.

We’ll see a split between frontier intelligence for high end tasks and work, and much cheaper models for high volume workloads that can sufficiently be peeled off to cheaper models. Frontier will still be far bigger than today because the use-cases will demand it, but the low-end will get quite a bit larger as well.

The big update here is that the layer that can efficiently route the workload to the right model will then become increasingly valuable since that becomes one of the new hard problems in AI agents. Agent orchestration that can cost optimize while still performing the task successfully will be in a strong position.

This is what the market got wrong about AI eating enterprise software. Building good software in the past was very hard. Yes, AI has made that a bit easier, though it’s still hard to build something that’s got good taste, differentiated, high quality, secure, and so on.

But nevertheless, that’s only one component of building a platform that enterprises rely on. The plurality of costs in most enterprise software companies is actually on GTM, because at scale most enterprise software categories are tough to break into and need a heavy amount of consultative selling and support for implementation and integration of solutions.

AI hasn’t reduced the need for that, and in many cases requires it even more now, as landscapes get even more busy and complicated for buyers to navigate through. If you make one thing cheaper and more abundant (development of software) then the new problem of discoverability and market differentiation (GTM) becomes the hardest part.

Box now has a markdown editor on the web. Full CLI support. Commenting. Full version history. Box Drive also lets you connect to any desktop client as a mounted drive, so you instantly work with all your files in Claude Cowork, Codex, Obsidian, Cursor, or any other app. https://t.co/3dJ5SBzhM5 https://t.co/WLUKegtiJ5

I think one reason that my Frontend Slides skill has grown so much organically is because slides are inherently social

People see these cool slides and always ask "how did you make it". And people tend to perceive those using HTML decks as more AI-native and AI-savvy

https://t.co/CiQpfibZn5

This is why we created https://t.co/NH6AWPWWkY

Educating people on how to use the AI tools has become a serious bottleneck https://t.co/x624uLrMkb

GBrain v0.42.30 can now give you a detailed summary of how your thinking has changed over time. https://t.co/lcj64PThft

A common misconception is that training data is low skill, grunt work - scan some notebooks, mine the internet, create labeled samples.

The data required to advance the model frontier is the opposite. Labs need training data for high-economic-value tasks. And most of these tasks outside of SWE have little documentation - it is complex, domain-specific knowledge built over the years, spanning legacy tools that don’t talk to each other.

That's why we have SWE agents and not knowledge work agents yet.

The companies creating this training data, such as Mercor, are doing extremely high-leverage, high-skill work.

Critical to moving AI forward. And deeply underappreciated.

Ok ChatGPT had a good one:

Wife: "I did, the loop is called marriage and this is your daily cron job."

My wife asked me to take out the trash.

I said, “You shouldn’t be prompting me anymore. You should be designing loops."

Finish the punch line of this dad joke please.

Links I mentioned:

Compound Engineering:
https://t.co/zo3cxxgMeD

Kun Chen's free tools:
https://t.co/aqEdDzZKPe

Given all the “loops” conversation, @Apple has the chance (and address) to do the funniest thing at WWDC.. https://t.co/HAhOpSW5Pn

The vibe shift from tokenmaxxing and token anxiety to tokenoptimizing in just a few weeks is wild 😅

Might be a hot take but I still believe companies should give copious amounts of token budget to employees to stay at the frontier and explore all the edges.

Otherwise, it’s way too easy to fall back to “doing the things how they have always been done”.