AIDataCenterGuy

@emollick @danshapiro It's funny, but when AI gives us a time frame, it's always in human time probably due to their training. In reality, a couple of hours and it's done.

Great little story from @danshapiro about how he asked a coding agent to fix the official webcam software from Canon that kept crashing. He woke up to a new, fully functional Rust webcam app that has worked ever since. danshapiro.com/blog/2026/03/t…

"Every software company in the world, needs to have an @openclaw strategy" - Jensen at @NVIDIAAI GTC Framing OpenClaw as one of the most important open source releases ever, they have announced NemoClaw - a reference platform for enterprise grade secure Openclaw, with OpenShell, Network boundaries, security baked in.

Get your dopamine from production, not consumption.

Oh and it works in all AI4-equipped cars, so your car can do office work for you when not driving. We’re also deploying millions of dedicated Digital Optimus units in the field at Superchargers where we have ~7 gigawatts of available power.

Just out: California's wealth tax proposal has a NEGATIVE $25B net present value. We project only ~$40B collected (vs. $100B claimed), with ~30% of the tax base already gone and lost income taxes wiping out gains. w/ Ben Jaros, @gregkkearney, John Doran & Matheus Cosso: papers.ssrn.com/sol3/papers.cf…

Founders who can write clearly are rare and worth paying attention to. Dan puts Dario in the same class as Bezos. The only signal that mattered early in Amazon was the '97 shareholder letter.

@plehoux Exactly the same for me.

I've never worked as much as since my work has been announced to be automated. And I've never had as much fun working. I feel like a teenager, everything is possible.

Bye bye Mac Mini. Welcome NVIDIA DGX Spark.

People like Tobi are flying because they’ve been training the muscle to make 100 high level ambiguous decisions a day for a decade Sudoku programmers who enjoy the calm satisfaction of solving logic puzzles are exhausted because they’ve never had to use this muscle so much before But mental muscles can be strengthened, it just takes time

@01Ananto Garage door opener apps, parking apps, third-party charging apps, Waze, etc.

Why do we need CarPlay when we already have the best infotainment system in any car?

The real story here is worse than a fumble. It’s a three-step own goal. January 9: Anthropic locks Claude Code OAuth tokens, killing every third-party tool that built on Claude subscriptions. OpenClaw, which recommended Claude Opus 4.5 as its default model, wakes up to a broken integration. No warning. No partner outreach. January 27: Anthropic’s legal team sends the cease-and-desist over “Clawdbot” sounding too similar to “Claude.” Steinberger complies at 5 AM on a Discord call. During the 10-second window where he releases the old GitHub and X handles, crypto scammers hijack both accounts and run a $16M pump-and-dump scheme. The chaos reflects on the entire Claude ecosystem. February 15: Steinberger announces he’s joining OpenAI. So Anthropic had the fastest-growing open source project in AI history (145K+ GitHub stars, 2 million visitors in a single week), built by a guy who sold his last company for ~€100M, whose tool literally recommended Claude as the default model to millions of new users. Their response was to cut off his API access and send lawyers. Steinberger spent last week in San Francisco meeting with every major lab. He explicitly said he could have built OpenClaw into a massive company but chose OpenAI because he wanted “the fastest way to bring this to everyone.” Meanwhile OpenClaw has already spread to China, with Baidu planning direct integration into its main app. This is a project that was essentially a free distribution channel for Claude. Millions of developers installing a tool that defaults to your model. The growth marketing team at Anthropic should have been sending gift baskets, not legal notices. Sam Altman just got handed an open-source agent framework with global distribution and a brilliant founder, because Anthropic’s legal department moved faster than their partnerships team.

CEO of Shopify @tobi is shipping more code than ever. 2024: 94 commits 2025: 833 commits 2026: 957 commits (in first 45 days of the year) Claude is turning CEOs back to builder mode.

@nim_alert @openclaw 😉 x.com/sama/status/20…

That’s an interesting definition of ‘not smart': @grok: Peter Steinberger has publicly confirmed receiving acquisition offers from both Meta and OpenAI. He discussed this in interviews, notably on the Lex Fridman Podcast (#491), where he mentioned these two stood out as the most interesting among talks with big labs (including even Microsoft/Satya Nadella).

Liquid cooling is quickly becoming the de-facto “go-to” cooling method for data centers. In this insightful QuickChat, we discuss what this means for growing data center demands, along with best practices going forward. #liquidcooling #datacenters bit.ly/3M2AE5Q

In two years NVIDIA made air cooling physically obsolete for high-density AI. GB300 racks pull 142 kW each. Air can't remove that kind of heat. The liquid cooling market went from $2B to over $5B since 2023. We sell water treatment chemicals. Data centers became our fastest-growing customer segment.

@amorriscode Great feature. However, there is no way to see which SSH server we connected to afterward in the UI.

SSH support is now available for Claude Code on desktop Connect to your remote machines and let Claude cook, TMUX optional.

@alightinastorm @RobiBuilds Exactly

@RobiBuilds the nuanced answer is that claude is smarter across all topics and a much more versatile model (for more than just code) gpt is a retard outside coding tasks, but oh man on coding it is very very reliable

wow gpt 5.3 codex is.. just wow canceling my claude subscription

Exactly what I've experienced.

@openclaw Oh, and I built that this morning while sipping my coffee in about 30 minutes with Claude Code Opus 4.6...

Built SmartRouter for @openclaw a tiny proxy that scores every prompt and routes it to the cheapest model that can handle it. 3 tiers, fully automatic: - "hi" → DeepSeek V3.2 — 20x cheaper than Opus - "check my emails" → Gemini 3 Flash — 10x cheaper than Opus - "refactor this auth module" → Opus 4.6 — full price, worth it It scores 9 dimensions (reasoning, code, complexity...) in <1ms and picks the right model. No config per task. No manual model selection. Result: ~80% of my requests don't need Opus. Heartbeats, tool calls, simple conversations — all routed to models that cost 10-20x less. Single file, zero dependencies, runs on localhost:4000.