There's a big war monument to Soviet soldiers near the Reichstag

You need to store a ton of water to get decently useful energy storage, so a lot of dams do this with huge reservoirs. Relies on having high yield renewables which aren't fully used off peak which you then store the energy from.

It's my understanding from reading the various documents there that the court found Russia's claims admissable, ie that the court had juristiction and the claims fall under this case rather than a new case. I'm not a lawyer so struggling to find where they discuss the merit of the claims themselves though. Can you elaborate on this further?

I work in tech and there are definitely a lot of engineers focused on efficient AI usage. Almost all of the best engineers I know are obsessed with efficient usage actually. Context lengths are limited and smaller contexts reduce hallucinations, so more efficient model usage also produces better outputs. The main challenge is getting high quality 'lean' context to describe your problem fully enough to the model, and there are many promising attempts at this for code problems which will eventually be reflected in other industries with huge knowledge bases that LLMs struggle with.

There is also tons of work on more efficient and smaller models. There are tradeoffs with smaller models in that they hallucinate more, can't generalise as effectively, and perform much worse. There is a technical limit on what they can model based on their size, though active research is still ongoing.

The whole premise of this strip is making fun of Trump by calling him gay (derogatory). It's fairly poor.

Have you tried 'microg 4 lineageos'? I've been using most of my bank apps without any issue through it even with strong integrity failing. Not ideal but it works best for me while minimising google's existence in my life better than stock android. It even has OTA updates much longer term than my oem. A major caveat for almost all phones is the firmware can basically never be free, and sim networks are designed for spying anyway.

None of them are a grift. They get used by financial traders for performing grifts, but the grift was and always has been stock market manipulation. AI is incredibly useful and many competent engineers now use it effectively. I expect the bubble of anthropic/openai will pop once local or on-prem models get better.

Jellyfin has largely the same, although I've never used Plex so maybe I'm wrong on that. It's a pain for the host to setup and get going with jellyfin for multiple users with auth (speaking from experience), but it's a do once have forever setup.

If you don't own something, eventually Capitalism will weaponise it against your soul. FOSS is owned in common, and self hosting is your own. Never understood why people liked plex.

An excellent deep dive on this is done in the "This podcast will kill you" podcast (episode 120). There are a few main theories on why but generally we don't really know much at all about its mode (or modes) of action.

Yep. We also don't really understand how paracetamol works beyond knowing it does work and has very few side effects.

In my day to day job as an engineer I use LLMs to help me do my work. They are tools just like anything else and enable me to work faster. I've even piloted an agent flow that takes our existing site (a huge behemoth of legacy and new services) and automatically performs accessibility testing across a variety of tools then gives us reports. We have actioned these to actively improve the experience of people using our product with screenreaders for instance (and have confirmed this from these very same people giving us feedback on the changes). The view that they are useless runs counter to reality and in my experience is usually from people who have deep emotional hate of technology. LLMs have incredible potential but as with anything under Capitalism they are used for horrors and their rollout is done with complete distain for human life and the environment. That is not special to AI.