Also a lot of vulnerabilities: https://app.opencve.io/cve/?vendor=ubiquiti

Technically they are, just a subset. But still a practical one, they're frequently used to produce executable files.

Depends on the GC. For some the amount of work is related just to live objects.

Fossil fuels use continues to increase and shows no signs of slowing down: https://ourworldindata.org/fossil-fuels

Instead of trying to push with a string (coming up with new non fossil energy sources) we need to ramp down fossils production using regulation, international agreements, emissions trading etc. Otherwise we'll just keep using both fossil and new energy sources.

That's a negative framing and while I'm not going to call it /wrong/, but world population is increasing so of course our energy usage is going to increase too.

If you scroll a little further down on your linked page, you will encounter another[1] graph, with renewable fraction of primary energy production and its steadily falling everywhere but the gulf states.

We're probably doing too little, too late, but my read is actually that we're moving in the right direction even if there is significant inertia to overcome before emissions actually start decreasing.

[1]: https://ourworldindata.org/grapher/fossil-fuels-share-energy

The US reversed course with the current administration and until things change with regards to that, there will be active resistance to renewable power projects. Because we need all the energy we can get and there's money to be made building out those projects there will still be progress, but it will be done in opposition to federal goals.

> we're moving in the right direction

many countries are, but the US is now moving in the opposite direction and because of its size it has a very large global impact

The PDF may be truncated, here's a better link:

https://sharedcurriculum.peteschwartz.net/files/Lovins_1976%...

Are there going to be self sustaining server factories in space? The hw needs updating every few years.

How do we bootstrap to the point where we eventually have factories in space?

If it were up to me, I would be researching and optimizing non-chemical propulsion, also green synthetic fuels for launches. Unfortunately, we will have to leave that for some firm other than SpaceX to develop.

Does making it "just work" here come with any caveats vs standard PG?

Getting there! Cross-shard writes do because of 2pc. Reads are eventually consistent.

Given that they implement connection pooling and sharding, I'm going to say "not at all".

You _could_ make that ACID, but it's not going to be faster than a single machine.

When I realized they're going to be largely powered by increased natural gas use in the USA, neatly combining with our biggest problem so far (the climate catastrophe).

Worst indeed, hardly anyone trades with them.

The Mac is very feeble compared to the big iron that the providers run so will be much lower performance. Also many companies would prefer engineers work on the domain problems instead of working on novel LLMs.

I meant “roll your own” LLM for use not build new ones.

The Mac Studio (and DGX Spark, for that matter) aren't running SOTA-level models by a large margin. Time is money, and waiting on these half-baked solutions is a waste of them both.

Especially concerning the Mac Studio, the GPU is far too weak for enterprise-scale context prefill. You'd need 2 or 4 Studios to process 250k contexts quickly, and even then you'd get bottlenecked by the relatively slow memory bandwidth during the decode stage. It is simply terrible hardware for quick or power efficient inference.