@bcherny Hm but you’ve said multiple times in the past you don’t use git worktrees and instead just have multiple repo clones. Did you mind change on them?
English
Kevin Moran
720 posts
Kevin Moran
@KevinTechTalk
Sr. Engineering Manager, . AR/VR enthusiast and daily AVP user. I bake sourdough and self-pack Biltong. My opinions are my own.
Bergabung Mart 2019
115 Mengikuti78 Pengikut
10/ Use git worktrees
Claude Code ships with deep support for git worktrees. Worktrees are essential for doing lots of parallel work in the same repository. I have dozens of Claudes running at all times, and this is how I do it.
Use claude -w to start a new session in a worktree, or hit the "worktree" checkbox in the Claude Desktop app.
For non-git VCS users, use the WorktreeCreate hook to add your own logic for worktree creation.
Learn more: x.com/bcherny/status…
English
I wanted to share a bunch of my favorite hidden and under-utilized features in Claude Code. I'll focus on the ones I use the most.
Here goes.
English
@badlogicgames to be fair, gpt finds issues in its own generated code too
English
i have real trust issues with opus. it's extremely weird. any time it suggest a thing, i switch to gpt and ask for a second opinion. in +60% of cases, gpt finds additional things opus just missed.
(not a gpt shill, love opus, but it's complicated)
English
@badlogicgames @steipete I mean if you another instance of Claude for a second opinion then you’d also likely get improvements. There’s already enough research showing LLMs verifying LLMs (regardless of model) lead to improved code quality
English
@mweinbach Whoa that must be new… never seen that before. Nice
English
@noahzweben @bcherny Why is this not just a configurable value?
English
Loops now run for up to 7 days instead of 3.
Let me know what you’re /looping on!
Noah Zweben@noahzweben
/loop 5m make sure this PR passes CI While loops for agents have dropped! code.claude.com/docs/en/schedu…
English
AI5 will punch far above its weight, because the entire Tesla AI software stack is designed to make maximally effective use of every circuit. We co-signed our AI software and hardware.
Bear in mind that AI5, while it can be used for training in data centers, is primarily optimized for AI edge compute in Optimus and Robotaxi.
There is still significant room for improvement. In the same half reticle and same process node, we think a single AI6 chip has the potential to match a dual SoC AI5.
English
Tesla’s forthcoming AI5 uses a half-reticle design, which is crucial for yield. A reticle defines the imaging area of a lithography machine, fitting two chips per shot effectively doubles yield.
This means the Tesla chip design team had to carefully manage die features, for instance dropping the older ISP (and classic GPU) to make room for more AI cores.
By contrast, NVIDIA’s Blackwell fills nearly a full reticle, making it a single-reticle design.
If Tesla hits its compute and efficiency targets with AI5 in this half-reticle format, it’s almost like cutting fab requirements in half.
And this has a big impact on Terafab, especially if it carries forward for AI6, AI7, etc.
phil beisel@pbeisel
Terafab may be the most essential vertical integration Tesla has ever undertaken— and it is truly non-optional. It will take years to build and will test even Elon’s speedrunning abilities to the limit, but that won’t stop him from trying. The breakthrough likely lies in overhauling the overall facility’s cleanroom model. By moving wafers in sealed pods with localized micro-environments, the fab no longer needs a monolithic ultra-clean space. Elon’s line about “eating cheeseburgers and smoking cigars” on the fab floor isn’t silly, it’s the practical reality of a radically simpler, cheaper, faster approach that could finally change the economics of chipmaking. This is all forced by the brutal “pinch” in chip supply. Tesla must produce on the order of 100–200 billion AI chips per year just to saturate its roadmap. That volume powers: FSD cars & Robotaxis (tens of millions of vehicles needing AI5 inference for near-perfect autonomy), Physical Optimus (scaling from thousands today to millions per year, each requiring AI5/AI6-level compute), Digital Optimus (the new xAI-Tesla software agents for digital/office automation, running massive inference clusters), Space-based data centers (AI7/Dojo3 orbital compute for GW-scale training and inference beyond Earth limits). AI5 delivers the ~10× leap for vehicles and early robots; AI6 shifts focus to Optimus + terrestrial DCs; AI7 goes orbital. No external foundry (TSMC, Samsung, etc.) can deliver that scale or timeline— hence the Terafab launch. Without it, the entire robotics + autonomy future hits a brick wall. Terafab isn’t optional; it’s the only way forward.
English
Ok so what local coding model should one use with a maxed out M5 Max MacBook Pro O_O
English
@X how is your iOS so bad? I’m trying to sign in and it literally tells me I need to use an official X app to sign in when it’s literally the X app…
English
@svpino Even anthropic seems to recommend iterm2 in their docs for like agent teams
English
With all the hype around OpenClaw and people building fast, I feel @steipete should like livestream a coding session of building new features. I know a lot of people who would benefit from some real agentic coding processes
English
@localghost It is quite bad… some people can do decent apps with electron but this is not one of them. Hope all is well!
English
Had my first *really cool* moment with @openclaw. I got a new travel mug that came with some chipped paint. I asked openclaw to look up the mug and see what the paint was made from. It told me but also told me to file a warranty claim with it since it would likely to be covered. I told it to do it for me!
It asked for the order number and shipping info and a picture of the damage and receipt and next thing you know it filed it and I got an email confirmation! Really cool.
English
@steipete So if we have OG clawdbot do we need to uninstall and reinstall openclaw? Just doing updates still keeps the clawdbot name around
English
at this point I usually prefer web apps. bugsappleloves.com
English
Wow! Elon announced $TSLA S and X will both be sunset this quarter! Absolutely 🤯
He’s already shutting down the car business before bringing up the new businesses.
Not sure what to think of this. But my thoughts are not positive!
As a Cybertruck owner I’m also concerned that he’s about to cancel the Cybertruck too.
English
@mweinbach @FredLambert I’ll take a large discount but only if they bring the Model Y L to the US. I don’t see how they can’t with the bigger X no longer being sold. They need something bigger than a standard Y
English
@KevinTechTalk @FredLambert i don't think they're capable, i think they make new camera units with the HW3 housing/connector but HW4 sensor/lens
either that or they'll wait until they're more cash flush, REALLY realize they can't do it, then offer refunds or large discounts on new cars
English
You know what this means folks? Tesla will have to disclose Cybertruck sales in "other models" now. 🤣😂😅
Electrek.co@ElectrekCo
Elon Musk kills Tesla Model S and Model X because of 'autonomy' electrek.co/2026/01/28/elo… by @fredlambert
English
This is pretty cool, bet this vehicle has a bright future
Tesla North America@tesla_na
Falcon wing doors can squeeze where your body can’t
English
Maybe but that’s only if you believe Elons comments that “the cameras are capable” on hw3. I’m a bit skeptical.
If he’s discontinuing these vehicles he should at least prioritize them for hw upgrades especially when those owners spent the most money to buy the car (and probably bought FSD back when it was 12k or 15k)
English
My bet is at some point they do it once AI5 is out and fab cost on the AI4 node is cheap enough, and they do a custom housing for AI4 chip to fit in an AI3 car with existing hardware. New computer using existing connectors and new cameras on existing connectors.
My guess is capable of unsupervised, but not as good as HW4 or HW5 native cars (but still good enough to fulfill that promise in full, just many many many years late)
English
@SawyerMerritt The only reason they are doing this is because no regulation will let them be on the road that way. Zoox does it because it’s a beta that isn’t allowed to scale up. They are doing this to buy themselves time to try to solve Unsupervised FSD before regulations pass
English
Elon Musk on Cybercab: "It does not have steering wheel or pedals. It either drives itself or it does not drive."
English