LostWanderer comments on Mistral AI's CEO says Europe has 2 years to stop becoming America's AI 'vassal state'

Mistral AI's CEO says Europe has 2 years to stop becoming America's AI 'vassal state' (www.businessinsider.com)

submitted 3 months ago by RockBottom@feddit.org to c/fuck_ai@lemmy.world

27 comments fedilink hide all child comments

cross-posted from: https://mander.xyz/post/52167452

you are viewing a single comment's thread
view the rest of the comments

[–] LostWanderer@fedia.io 65 points 3 months ago (3 children)

Man, these fuckers are so goddamn weird...Europe could just not use LLMs and instead pour resources into solidifying their access to better technical infrastructure, European made services, and parts for more fruitful endeavors. The USA is on a doomed route that will fuck so much shit up for a very long time. It's the most ill advised and thoughtless path that I've ever seen. Over a deeply flawed, easily exploited, and profoundly unprofitable technology...It baffles me that European leaders as a whole don't appraise it as a failed venture and invest in better things.

hideshow 6 child comments

[–] njordomir@lemmy.world 6 points 3 months ago*

Europe, especially the EU itself loves to do at least some of the stupid shit America does, just a few years behind our schedule. We are global leaders... at doing stupid shit.

parent

[–] Semi_Hemi_Demigod@lemmy.world 2 points 3 months ago* (last edited 3 months ago) (3 children)

There’s a distinct lack of competition from Europe when it comes to hyperscalers like AWS or GCP. If you ask me they’re at least two generations behind the US in tech.

ETA: I would love for Europe to be at parity to the US and China, but they’re not. My hope is that open source will help them figure this out quickly.

parent

hideshow 6 child comments

[–] LostWanderer@fedia.io 11 points 3 months ago (2 children)

Yeah, that's because Europe was moving slower with intention...As they cared about the impact of the technology on their citizens, at least this is how I see it as an American. I always felt that the USA moved way too quickly and without enough regulation to correctly constrain corporations or protect their citizens from said corporations.

parent

hideshow 4 child comments

[–] jacksilver@lemmy.world 8 points 3 months ago (1 child)

You can't both "move slower with intention" and still consume the "move fast and break things" products.

If Europe is going to allow the big corporations to chip away at their corporations, then they'll get the worst of both worlds; reliant on foreign technology with no in house capabilities.

parent

hideshow 2 child comments

[–] LostWanderer@fedia.io 8 points 3 months ago (1 child)

That is why Europe needs to invest in true tech sovereignty and not LLMs because LLMs are honestly a dead-end that will only produce useless results. They need to shore up open source projects, OS, infrastructure, sustainable manufacturing, and companies that can provide useful/productive services. I also think that figuring out how to extend the lifetime of already existing tech in a fruitful way...Would be a great route too, permacomputing could be a good avenue to pursue. To cut down reliance on producing a steady stream of components. Instead, creating parts with long-term usability and repairability in mind.

This will reduce and remove the reliance on USA technology, making them powerless in the European space.

parent

hideshow 2 child comments

[–] njordomir@lemmy.world 4 points 3 months ago* (1 child)

Yeah, this is a distraction to make them waste time. In 2 years they could be 100% completely weaned off of Microsoft SaaS if they stepped strategically. That's much more important than generating pics of big titty anime girls and 15 second videos of talking animals.

Interesting idea on the perma-computing. Imagine if they made modulate repairable gear that could survive a solar flare. Even if it was slow and clunky 90s level CPU tech, critical systems remaining functional when everyone else's shit gets fried would be a massive win.

parent

hideshow 2 child comments

[–] LostWanderer@fedia.io 1 point 3 months ago

European leaders if they play their cards right, they could wean themselves off of Microsoft SaaS in a reasonable timeline and have a sturdy system to rely, I do agree with this statement. The problem is that they've been hemming and hawing for way too fucking long on the matter. Not planning a strategic exit by bolstering open source tech and OSes that aren't controlled by shady US Technofascists. They are finally acting after years of deliberation and they are taking a slow approach this year by investing in open source and first transitioning to open source software/document formats. I can only hope that their plan succeeds, American Technofascists need to fail to hold power in Europe, like they have in China as well. As for the big titty anime girls...Anything has higher priority to me than that in my opinion (I prefer hand-drawn or animated husbandos myself, nothing made using a LLM slop generator). There is no justification of the technology that would pass muster in my opinion, as it is based on theft of copyrighted material, human made creative works, and actively is detrimental to the environment.

As for the permacomputing; it wouldn't even necessarily have to be clunky 90's level tech. You could use some modern tech as a base and figure out how to optimize the lifespan of the component and the ability to repair said component. If you also bear in mind that you'd design an OS in tandem with hardware...The experience would be incredibly optimized for daily usage. You could make interfaces that would make everyday people feel comfortable, they could still be reasonably reliable, critical infrastructure would naturally be handled differently. As withstanding a solar flare and being repaired after that disastrous event would be quite handy. The problem with modern tech is that we've let manufacturer's build in planned obsolescence into their products, that planned short lifespan needs to be removed and replaced with long term support!

parent

[–] Semi_Hemi_Demigod@lemmy.world 0 points 3 months ago (2 children)

Yep. But that caution had a trade off and they’re now having to play catch-up

parent

hideshow 4 child comments

[–] LostWanderer@fedia.io 5 points 3 months ago (1 child)

Eh, I still take Europe's situation over the tech hellscape of the USA right now...If I'm totally honest. It's so cursed here.

parent

hideshow 2 child comments

[–] Semi_Hemi_Demigod@lemmy.world 3 points 3 months ago (1 child)

Same. Even as an American I still find Europe the largest hope for democracy and the rule of law.

And it’s not a very big hope.

parent

hideshow 2 child comments

[–] LostWanderer@fedia.io 2 points 3 months ago

Still a better hope than what we have here, as long as we can make these Fascist fucks that are in power in the USA realize they can't win...We're going to be good. It will just be a long, sustained fight, and they are going to try to destroy our hope constantly. Other countries have won this fight, we fucking can as well...Then we need to address the root cause, as Trump and the technofascists are merely a symptom that manifests.

parent

[–] njordomir@lemmy.world 3 points 3 months ago (1 child)

You could look at it that way. When I look at it, I see the leader walking down a bad path and those trailing behind still have the chance to take a different path. Just because the US ran "ahead" doesn't mean the world should follow them off a cliff, even if they are screaming "wow, these AI waifus are so cool". Much of AI isn't explicitly progress, just change that has been framed as progress through the use of propaganda. If it's a regressive change, not moving fast could actually result in the EU getting ahead, like a reckless rock climber picking a bad line and falling back down to the last anchor while a careful climber avoids that route.

parent

hideshow 2 child comments

[–] Semi_Hemi_Demigod@lemmy.world 3 points 3 months ago* (last edited 3 months ago)

Yes. Awful things are awful. Humans suck. Etc.

parent

[–] TheJesusaurus@piefed.ca 2 points 3 months ago

This is an insane take lol.

parent

[–] Zos_Kia@jlai.lu 1 point 3 months ago*

We have good European clouds, what are they missing in your opinion? In my experience they're pretty cheap and reliable.

Interestingly I've been hosting my shit on EU servers for years with very little issues, while all the US services I use have appalling QoS (Google APIs, and hyperscaler hosted stuff such as N8N and other SaaS products).

parent

[–] Zerush@lemmy.ml -3 points 3 months ago (2 children)

AI is needed in science, medicine, meterology and industry, eg.investigations by the CERN are not possible without AI, some meta-materials which we use wouldn't exist without AI. AI itself isn't the problem, but as hype inbuild even in a Fridge, it's use by any idiot, tankie and big (US) corps with selfish intentions. Europe has good own AI alternatives, eg. the one used, among others, by the CERN, the OpenSource Apertus.

parent

hideshow 4 child comments

[–] azolus@slrpnk.net 15 points 3 months ago (1 child)

Part of the problem is people generally confusing Machine Learning and other actually useful forms of AI with the generative LLM slop, biometric mass surveillance and autonomous weapon systems these tech bros are pushing for. So no, we don't need that kind of "AI".

parent

hideshow 2 child comments

[–] Zerush@lemmy.ml -2 points 3 months ago (1 child)

Agree. Well LLM is inherent to make possible an interaction with the AI and asociate machine learning. But certainly it isn't the same using a biased chatbot with an LLM to analyze, interprete and summarize a given text or data set. Bad is when the generative AI invent datas when it don't figure in his lenguage model causing this slops. This often ocurre in alround AI which pretend to do everything (text, image,......), not so in AI specialized in certain tasks. Like a swiss army knife never can substitute a normal knife suissors, saw, screwdriver......, it will always only be an reduced emergency solution.

parent

hideshow 2 child comments

[–] jaredwhite@humansare.social 12 points 3 months ago (1 child)

You don't need an LLM to interact with machine learning algorithms.

In fact, one might argue one should not use an LLM to interact with machine learning algorithms. 😆

parent

hideshow 2 child comments

[–] Zos_Kia@jlai.lu 0 points 3 months ago

That's a very exotic take. Language (semantic embeddings) is a good mediator for a lot of tasks, which makes it easier for ML models to generalize. The intersection of old school ML and LLMs is a fascinating place, I've never seen it dismissed like that. Do you have any reading you'd recommend on the subject?

parent

[–] ZDL@lazysoci.al 6 points 3 months ago

Just as a professional communication tip here: When the general public is speaking of "AI" right now, they are referring to degenerative AI, not to the older waves of AI hype and collapse.

The older waves have found their niches and perform admirably in them. (My phone's camera was "AI" before LLMs became a thing, for example.) But that's not what anybody but a pedant (who is only barely better than a pedarast) means when they say "AI" outside of very specific technological communication contexts.

parent