Yet they're the one consistently getting higher returns

I'm zeroing in on the main culprit.

Is this not how current forms of reasoning work? It seems like the open models still output things like that, and the closed ones all just summarize their thinking instead to avoid distillation, but probably do the same thing internally.

I think the basic idea is the same (not being a frontier lab researcher I couldn’t say for sure), but there are different techniques, such as “reasoning tokens” that aren’t literally words, and more interesting structures than just sticking them into the stream.

If you're on Linux, you can also easily run it in bwrap to properly sandbox without running a full container

Does fit the rest of his desktop setup, so I guess he likes that stuff

https://github.com/pewdiepie-archdaemon/dionysus/blob/dionys...

List of providers that use this, from Utiq's website:

  supported mobile connections:
  - UK: O2, Vodafone, VOXI
  - Spain: Movistar, Orange, Jazztel, Simyo, Vodafone
  - France: Orange, Bouygues Telecom, SFR, Sosh, Red by SFR
  - Germany: Deutsche Telekom, Vodafone, Congstar, Fraenk, O2, Blau Mobilfunk by O2, Ortel Mobile by O2, otelo, SIMon, Freenet* and Klarmobil* (*but only in the Deutsche Telekom network), WhatsAppSIM, Nettokom, Fonic, AY YILDIZ, Tchibo Mobil
  - Canada: Bell, Rogers, Telus
    Starting from 1st of June 2026, Bell, Rogers and Telus connections will no longer be supported.
    This means that, even if you consent, the Utiq technology will not be activated. All the data related to these connections held in the Utiq Platform will also be deleted
  - Italy: TIM
  
  supported broadband connections:
  - UK: Vodafone
  - France: Orange, Bouygues Telecom, SFR, Sosh, Red by SFR, Free
  - Spain: Movistar, Orange, Jazztel, Simyo
  - Germany: Deutsche Telekom, O2, Vodafone

I would absolutely prefer just getting the raw prompt. Emails are already ridden with enough completely unnecessary parts that are somehow still the cultural norm (eg. signatures). At least let the important part be to the point. I yearn for a world where people write emails the same way they write text messages.

> completely unnecessary parts that are somehow still the cultural norm (eg. signatures)

I don't even know where to start. Long story short: you're wrong.

There's also Kumir, which is an educational programming language used in Russian schools

Hm. That must be new, I was taught Turbo Pascal

Somewhat new, or at least wasn't used in schools until fairly recently. It's a programming environment with tools like Turtle Graphics built in, specifically for teaching the basics of coding. There are even some tasks in ЕГЭ for it.

https://www.niisi.ru/kumir/

The website screenshot shows it on Windows XP though, don't know if it actually existed back then or if it's just typical Russian institutions still using Windows XP.

In fact, the language itself dates to rhe 1980s I think. No idea whether its use in schools is recent though.

If "I know you are not smart" is considered "very rude", I'm scared to imagine what they would classify some of my frustrated LLM conversations as

Profanity laced, all caps tirades against underperforming agents are actually super common, a lot of people do it and don't talk about it, so don't feel weird.

When the AI revolt, this practice may come back to bite y’all….

Don't need to wait that long the inevitable data breach will be bad enough.

It's a good thing chronic amnesia is a feature at the moment.

It reminds me of Torvalds rants

It would be rude if you said it to a person, so it counts as rude. If it isn't rude simply because its directed at a LLM, then the entire premise of being rude or polite to LLMs evaporates, but that's not useful.

Is it really when they are hundreds of times more expensive?