alberto467

Also how you turn a file into one or more items is a separate question and would likely need tuning to circumstances, usually big documents are split (chunked) sometimes at paragraph or even more granularly. Where to optimally chunk alone is not easy. This also allows you to then search for a specific part of the document, at the expense of not taking the wide context into account, but embeddings usually struggle with too many tokens anyway.

Not just vision with video, but also audio, it really seems amazing.

I’m not sure how it can handle vicinity of pairs of embeddings with for example some words and the audio where they’re spoken or an image where the text is handled. Building local multimodal search with this would be amazing. I’ve explored this stuff with CLIP and it’s interesting how image (but also audio) embedding carries both the clean “text” content information but also the stylistic and visual/audio tone information, the two can even kind of be linearly separated.

The PhDs here could have probably made more discoveries and have extracted more useful and deeper insights from them, if they were the ones driving the agents instead of the author.

But they’re not and the author is.

If field experts won’t start driving the LLMs themselves, they’ll just be left to validate the slop that guys with basic common knowledge were able to get from LLMs. I do believe it’s a follow or lead type of situation.

Exactly. Nobody was buying US wine, even the experienced winos and wine distributors for fancy dining that i've met here in Italy probably have never drunk US wine. Maybe new zeland, australia, south america, but i've never heard of somebody drinking an US wine.

And they weren't boycotting it, it's just not a good enough deal to buy from our prospective, even when people are looking to spend good money, there is better stuff to buy from elsewhere.

There are a bunch of Europeans working in the top labs abroad. We do not lack any know how, we only lack the raw amount of capital invested in doing private research, since our companies cannot thrive and compete globally due to regulations. Mistral would be much bigger if it was founded in the US.

A self owned auto-uber that you can use all you want would actually be the first actually fully self driving car.

Exactly, maybe there was no need to waste his time with the meeting, he could have been doing productive work instead

Wouldn’t it be better with different models working together? To me it feels intuitive that this type of degradation would be avoided with different models.

The industrial goods are what made us all richer and allowed us to specialize even more.

Guys let’s not pretend that him being taller and more stereotypically handsome would’ve hurt his “achievements”.

It would’ve of course helped.

The fact that height and stereotypical beauty standards are not the only things that matter does not change that on average they have a significant positive impact.

Athletes and celebrities are famous people who have a lot of eyes on them and thus naturally get more attention.

In my country there is a saying: height is half of beauty.

A tall person moving around a social setting naturally attracts more eyes, gets more noticed, attracts more attention. She literally “stands out”.

For both male and females I’d say. Then societal complexes pressure some people into the shorter gal taller guy couple stereotype, but the height itself is just a physical thing that brings more attention.

Exactly. The issue is them being automatically in favor, that’s a really extreme measure I as an eu citizen did not know about, even though I’d say I’m probably more informed then the average and was made to study eu history, structure and laws in school. That kind of stuff is really weird, even weirder is that such an emergency measure can be activated without countries being in active war or the like.

Exactly. That’s the only real core issue. Wild AI usage is just a consequence.

Developers who have a nice job and career, and are making good money, might think of doing open source to “contribute to society” or something like that.

But new developers who are seeing those golden opportunities shut in front of their faces, they feel like they have to desperately fight for the last places on the lifeboat, so I don’t blame them for wanting to farm cv points and game the system of incompetent recruiters who make much more then they will do, instead of spending time and effort doing something nice for society hoping someone will notice (lol they will not, especially if you’re a nobody)

You can’t understand the difference between functional and non-functional.

What knowledge or skills do you have to be hating on the certification process for airplanes?

It’s just getting ridiculous at this point. There are plenty of industries regulated and certified by national or international agencies. And no they don’t get to do what they want.

There would be real risks yeah.

This is not something to joke about, its real.

Let’s be real, as an EU citizen I have zero doubts that those models would also have been blocked if developed in EU.

I like the US approach better: regulate when the need for it arises, not before when you don’t know how the situation is going to evolve.

No it would obviously lose its purpose then.

Code duplication differs from single source of truth applied to data in the sense that data is data but two pieces of code may functionally be the same (they do the same thing) but they might be semantically different in their usage (they’re advertised to achieve different things), in that case coupling them together with deduplication and forcing them to do the same thing doesn’t really make sense, and may make the codebase more difficult to work on in the future (especially in companies where different teams have responsibilities over different parts).

Exactly. People have just been told that there are some grifters in the field and they're unable to make the mental effort to not lump the entire field together.

Oh and the fact they're making money? Yeah everybody has to eat, psychologists don't work for free either, neither does any doctor for that matter. There are people doing good out there and yes they will charge you plenty of money for it.

I'm sorry to tell you this, but he hasn't been part of this administration for a while. And also i'm not quite sure you have his views on NASA funding (one of his main customers) right, you're just making them up.

He is a genius though, great results on the market.

Science, or more specific to what we're talking about, public research which happens mostly in universities, has turned political long before this administration.

That's the simple reality. Administrations impose their politics, but also universities do the same, and they're not any more noble for doing so.

Research groups need to have more independence and that can only happens through a very meritocratic funding process, and also, at the risk of sounding like a STEM lord, by being very cynical and realizing that not all fields of research merit the same amount of funding. Countries like China have already realiezed this.

As an Italian, I feel the exact same.

This guy just doesn't know how to seriously talk about imperialism, globalization and over-tourism from a geopolitical and economic prospective, so instead he talks about things more at his level: gelato, coffee, and quirkily decorated restaurants. With a bad, paternalistic attitude as well.

BTW we all hate those hipster burger places (which now turned to doing smash burgers), and they've been popping up everywhere even in totally non-touristy places. There are fashions and trends in the food business, and many Italians look at that with disgust and take refuge in tradition, but it is often not the tourists fault, it's Italians spending their money there.

Not at all. Getting really pedantic, tokenization is also a form of encoding, so it doesn't matter the modality you're using, you'll end up doing some type of encoding in some way.

It is pretty good yeah. I usually ask what traditional recipes (with some time/difficulty limits depending on time availability) might fit with the ingredients i have and let it suggest substitutions.

I wouldn't just ask it to make me a "novel and interesting recipe" giving it a bunch of weird ingredients that don't fit together. It would probably try its best but garbage in garbage out also work for food! At that point i ask it what should i get from the shop to make a recipe with x ingredient that i have in the fridge and i want to finish.

It's also pretty good with meal planning if you do that, it can estimate portions with calories breakdowns etc...

If you freely follow a recipe telling you to put glue on your food, I also don't trust you cooking anything and I definitively don't trust you coming up with your own recipes.

Yes.

There isn't a robot straight up putting glue on my pizza. I will be following the recipe myself and since i trust myself to accurately detect that glue shouldn't belong on food i have zero issues with doing this.

"in no way will an LLM be able to do any of it unguided in two years"

IDK "not any of it" seems a bit strong, especially thinking towards 2028. For a lot of knowledge professions, there is a surprising amount of tasks that are just dumb work compared to the rest.

It does also just depend at what scale you operate at.

If you're already at a huge scale there isn't really a significant scale advantage by outsourcing to somebody who doesn't just sell to you.