I only have pretty tame actions workflows and I have had a hard time replicating simple set ups with this. I can't imagine a company with more complicated setups.
What I wish is github codespaces could just do this out of the box, at least for a specific action/runner.
The problem with private address ranges is that everyone thinks they're available. In a large enough enterprise you're bound to have conflicts. They usually pop up at the most inconvenient time and suddenly you're cosplaying ARIN in your IT department.
I'm okay with writing developer docs in the form of agent instructions, those are useful for humans too. If they start to get oddly specific or sound mental, then it's obviously the tool at fault.
> In the 12 months to June 2025, wind and solar (2,073 TWh) generated more electricity than all other clean sources (nuclear, hydro and bioenergy) combined (1,936 TWh). Just four years ago, wind and solar generated half as much electricity as other clean sources combined.
So, both types generated approximately the same amount of power and it still isn't enough, one type cannot replace the other, they complement each other, that's why China is building more of each type, they know what they're doing.
Those are not really great construction examples, are they? Both projects took 15+ years to complete with huge cost overruns. And for those two "successful" projects, you can find 2 or 3 that failed.
The Finnish reactor had one delay because the concrete used for the containment building wasn't of the 'nuclear grade'. That's why those regulations thankfully exist.
Building more will help though. This whole thread started about how we had lost important knowledge
Specialized EV tires are also optimized for drag and noise, wear is just a factor. Anecdotally speaking I'm at over 80k km on a set of EV tires with at least 20k more to go. The issue has more to do with driving style and engine power than any other factors.
100%. Tire technology is a real thing. Tires have advanced a ton in the last 10 years. But driving style is the biggest thing. Some people can only get 10k miles out of a set of tires, while others with the same car and tires get over 40k.
But they do care about tire wear a lot, they know the acceptable wear life for the class. A couple years ago I bought a set of Pirelli tires that were ~50% off because they were an older version; hoping I’d get some benefit. Unfortunately they had half the life and were a bit worse in every way than the newer tires I had before and after.
Some tires are going to wear fast no matter what. I had some Pirelli PZero summer tires that I could never get more than 15k out of regardless of how I drove. The tire compound was very soft and sticky.
If you have something like really high performance tires, I recommend just using them. The grip is always there and you are always paying for it. As long as you aren't losing traction constantly, the difference is negligible in my experience.
No fundamental reason not to power them with renewables, either off-grid or with a small capacity grid connection. The argument that they need to run at full load 24x7 sounds more like a business requirement than a technical one. LLMs in particular with their stateless nature seem like an ideal candidate for global distribution.
The capital cost of the AI hardware is high and it depreciates quickly, being worthless in 5 years (or less). To make a profit, it needs to be run 24/7.
It’s sort of like how airlines like to fly their airplanes as much as possible.
The final analysis is still pending, afaik. In any case maintaining grid stability is a good problem to have and likely much easier to solve than generation. Worst case, you spin some flywheels to get inertia.
I'd really like to see the math on this one. It implies that building wind and solar on Earth is somehow worse than building it in space _and_ moving the data center there? It's not just counter-intuitive, it's bonkers.
you need to look at numbers instead of intuition - intuition naturally gets us wrong when we deal with unfamiliar things like space.
A ready 10 ton house will cost $500K-1M to place in a well developed area - the cost of land itself, communications, permits, also delays and unpredictability of process, etc. Add to that yearly taxes. Add tremendous electricity costs in case of the datacenter. And all those costs have and will be growing.
Launching 10 tons on Starship - sub-$1M and that cost will drop down to about $100K-200K.
> It implies that building wind and solar on Earth is somehow worse than building it in space _and_ moving the data center there?
reply