Sonori

This is also how passive RFID tags work, the tag harvests just enough energy from the scanning frequency to boot up a microchip and respond with its ID number.

If you have a hypervisor in a home lab, which it turns out OP does not, odds are good your already running a DNS filter/ server, DHCP server, AD domain, etc so the whole network is down anyway during hypervisor restarts if you don’t have HA setup.

OpenAI’s algorithm like all LLM’s is designed to give you the next most likely word in a sentence based on what most frequently came next in its training data. Their main strategy has actually been to use a older and simpler transformer algorithm, and to just vastly increase the scrapped text content and recently bias with each new release.

I would argue that any system that works by stringing sudorandom words together based on how often they appear in its input sources is not going to be able to do anything but generate bullshit, albeit bullshit that may happen to be correct by pure accident when it’s near directly quoting said input sources.

If you have an always on server, you can always run Opnsense as a vm.

I’m amazed it held out for so long. Small stacks and getting people used to useing your tool sounds like a good lon green strategy, and Boradcom doesn’t do those.

Why would a program that outputs sudorandom strings based on how often they appeared after the string before excel at basic physics, even if it did have most of the text side of the internet fed into it to determine how often string b follows string a? It’s a miracle of programing and self reinforcement that it can form a sentence at all.

There are limits to battery production, especially on short time frames. If your expecting every nation to try and deal with storing days of electricity production to cover for a rainy week your going to run out of easily accessed raw materials such as lithium.

You need either reliable generation, absurd quantities of undersea cables, or scalable storage. The only practical storage tech we’ve seen is hydro, and there are limited places to flood in order to construct built massive resivors, and it has far worse timelines and costs than nuclear, so that means on demand generation.

In this category we have nuclear, and location dependent options like hydro and geothermal. All of these are about as expensive, but output constant or at will power.

If you let the market decide, it’s going to do what its already decided to do, which is cheap solar plus cheap gas and coal. If you ban gas and coal, then it will be cheap solar and batteries for nations that can afford them and all gas and coal for the poorer nations that can’t afford the batteries.

Leave the batteries for applications like transport and smaller grids that need them instead of brute forcing them into places where they don’t fit like long term grid storage.

Finally, though this is the most minor, nuclear is by far the winner in local environmental impact, as it lacks the land use and habitat distruction requirements of solar, wind, and hydro. It’s also location agnostic, and unlike batteries gets cheaper and faster as it scales.

Counterpoint, wind and especially solar are now so cheap that the average grid scale solar plant turns a profit in 10 years and continues that profit for the next twenty plus. It’s cheaper per watt than gas and especially coal.

Perhaps govement subsidies should instead go to the less profitable 24 hour sources of power needed to fill the gaps, like hydro, geothermal, and nuclear, instead of just making already profitable investments a bit sweeter. There is a reason why well managed grids use a diverse set of sources, so unexpected shortages in one tech don’t limit the whole system.