I think a file delete just removing an adress and not the actual data is common to all OSes. That’s why to safely erase data from a disk it is recommended to fully overwrite the disk with random data, potentially multiple times.

They removed the paw, they are worse than Chrome! I’m going to use an alternative version that cannot exist without the main project to teach them a lesson!

Je pense que c’est une référence à la vidéaste scientifique Valentine Delattre qui fait des vidéos déjantées de vulgarisation de la géologie. https://youtube.com/@Sciencedecomptoir Elle a une sidekick qui est sa soeur jumelle morte imaginaire Fiona qui fait des featuring de temps en temps.

As well as unofficial “tourists”.

Ou Fiona ?

French Qwant and German Ecosia are building a EU based index together, but it will probably take time to be competitive compared to Google. https://www.eu-searchperspective.com/

The bangs are the addictive feature for me. You start by typing !w lemmy in the URL bar, with DDG already configured as the default search engine, and DDG will directly serve the search result for this term from Wikipedia. There are hundreds of them like !yt for YouTube or !gm for Google Maps, and the acronyms are intuitive. https://duckduckgo.com/bangs

I’ve been using it for a couple of year now and I only double check Google maybe 1% of the time. And most of those times, I also can’t find what I am looking for on Google. I also can’t go back after getting used to the ! commands.

They are socially conservative, but they have been fairly consistent at asking people not to kill each other for the past decades.

Ecosia and Qwant are building together an EU based index. Maybe then they will become actually independent alternatives to American giants. It will take a while.

There’s better integration with all sorts of other sources of truth beyond the LLM training, which makes it seem smarter.

Is there any other useful edtech than the FOSS app Anki?

L’informatique quantique a un vrai potentiel, la question c’est plus quand, et qui va réussir à développer les premiers produits industrialisables.
Il y a une bonne dizaine de prix Nobel de physique ces dernières décennies (dont les français Serge Haroche 2012, Alain Aspect 2022 et Michel Devoret 2025) qui attestent que c’est un sujet très chaud.

On parlait déjà de première révolution quantique pour les semi-conducteurs qui sont basés sur des phénomènes produits par un grand nombre de particules quantiques (physique quantique + grande statistique). La deuxième révolution quantique c’est d’arriver à exploiter les phénomènes quantiques qui se manifestent à l’échelle d’un petit nombre de particules (ex: superposition de plusieurs états sur une particule). Ces phénomènes ont des applications théoriques qui peuvent, entre autre, démultiplier la puissance de calcul, pour des problèmes spécifiques, par rapport à l’informatique actuelle (ex: exploration ultra-rapide des combinaisons pour créer des molécules ou pour casser un chiffrement).

La difficulté c’est que les états quantiques sont extrêmement fragiles, la moindre interaction avec l’extérieur et le système perd ses propriétés quantiques et devient inintéressant. Donc il y a une recherche appliqué intense pour franchir cette difficulté et rendre ces systèmes industrialisable. Je pense que les premiers à sortir un bon produit pourrait ridiculiser le boom actuel de l’IA. Plus de détails: https://www.nobelprize.org/stories/high-hopes-for-quantum-technologies/

It already has a social impact on the people’s whose intellectual property was pirated, the employees traumatized by nsfw filtering tasks and the reduction of white collar junior recruitment (probably non junior too?). I know some people think it’s just a bubble, companies are waiting to see what happens and the job market will recover. I am doubtful of it.
My tech company is pushing us to use it so they provide the top tools. From what I have observed, I have little doubt it will replace a lot of the designing, engineering, coding and communication time. Yes, you will still need some knowledgeable person to guide and review, but less than before. Similarly to how you need less people to build a car today than you did in the 50’, and even less for an electric car, because so much more is automated.
So far automation and the internet did create more better paid jobs than it destroyed, maybe it will happen with “AI” too, but I am skeptical.
Finally, in my opinion, UBI and work time reduction with equivalent quality of life is a desirable future.

AI could make sense one day if it comes with UBI to compensate its social impact and if it optimizes enough processes to compensate for its ecological footprint. We are far* from this and there’s nowhere enough political pressure to make it happen.

Gentle reminders at my work the past months: sign this company policy change that unilaterally changed your benefits with no upside for you, please. (Or you will be in trouble)

Anyone has a link to the original? Couldn’t find it with a basic search.

Bingo

I was reading per chip x employee, and I was like damn, that’s a lot.

Maybe Anthropic somehow attracted more politically conscious people compared to OpenAI, and it shows in the training?