AI engineer: “I built this model.”
Meta: “Excellent. We no longer need you.”
Meta is axing 600 roles across its AI division
491–500 of 506 posts
Re: Meta is axing 600 roles across its AI division
#492Earlier quoted context omitted.
I'm seeing a lot of frustration at the leadership level about product velocity- and much of the frustration is pointed at internal gatekeepers who mainly seem to say no to product releases. My leadership is currently promoting "better to ask forgiveness", or put another way: "a bias towards action". There are definitely limits on this, but it's been helpful when dealing with various internal negotiations. I don't spe…
Big tech is suffering from the incumbents disease. What worked well for extracting profits from stable cash cows doesn't work in fields that are moving rapidly. Google et al. were at one point pinnacle technologies too, but this was 20 years ago. Everyone who knew how to work in that environment has moved on or moved up. Were I the CEO of a company like that I'd reduce headcount in the legacy orgs, transition them to…
I'm being nice about this and assuming Google (which also owns Youtube, and it shows, in their decisions about content) is just run by a bunch of naive people and isn't directly controlled by three letter agencies - they have to take a stronger stance against censorship/selective information
"Oh, but that doesn't happen" - it does, Goog results have been manipulated before to the extent that probably can't be attributed purely to SEO. Youtube removed tons of "covid misinformation" about things we all know now to be true
Re: Meta is axing 600 roles across its AI division
#493Earlier quoted context omitted.
I don't think it does. The models have increased greatly in capabilities, but the competitors have simply kept up, and it's not apparently that they won't continue to do that. Furthermore, the breakthroughs-- i.e. fundamentally better models, can happen anywhere where people can and do try out new architectures, and that can happen in surprisingly small places. It's mostly about culture and being willing to experimen…
You don't think it does because what you described is only the optimistic take on how much farther LLMs will be able to advance :). The pessimists would look at previous "AI" and point out each new approach quickly rises to prominence and then drastically tapers off in how much more can be squeezed out of it. I'm somewhere in the middle. I think there is more to squeeze out of LLMs still, but not nearly the kind of g…
I'm not sure I think the progress is about multimodality. After all, Mistral's approach hasn't involved multimodality, and they've kept up.
An efficient model, then data curation, then post-training. Where things are slowing down is of course necessary to know to be efficient, at least in the short term competition.
Re: Meta is axing 600 roles across its AI division
#494Earlier quoted context omitted.
> sometimes they manage to meld it into one goal, because money is power. Money is a measure of power, but it is not in fact power.
Power is a measure of money, but it is not in fact money. See https://hbr.org/2008/02/the-founders-dilemma or the fact that John D. Rockefeller was furious that Standard Oil got split up despite the stock going up and making him much richer. It's not so clear what motivates the very rich. If I doubled my income I might go on a really fancy vacation and get that Olympus 4/3 body I've been looking at and the fast 70-30…
Wealth is cardinal. Power is ordinal.
Wealth is something that is counted and accumulates (or decrements).
Power is ranking. If you double your wealth, but maintain (or reduce) your power ranking, net-net you've lost power.
There are other elements at play. Discretionary wealth and power also matter. If you're in a position where all your wealth and/or income are spoken for (e.g., a business with high cash-flow but also high current expenses such as labour, materials, services, rents, etc.; or a governmental unit with high mandated spending / entitlements), then a numerically high budget still entails little actual discretionary power. Similarly, an entity with immense resources and preeminent ranking but where most or all options are already spoken for, where there are no good discretionary options, has nominal power but little actual power.
A classic example of the latter is a regime which embarks on a military misadventure only to find that it can pour in vast amounts of blood and capital for little actual return, ending up bogged down in a quagmire: Vietnam, Afghanistan (multiple instances), the Western Front (WWI), Gallipoli (WWI), winter invasions of Russia (Napoleon, Hitler/Barbarossa), the Charge of the Light Brigade, Waterloo, Agincourt, the Spanish Armada, etc.
Re: Meta is axing 600 roles across its AI division
#495Earlier quoted context omitted.
There's a segment of people convinced that leadership must somehow be able to perfectly predict the future or they're incompetent losers, like running a business is somehow the easy part of capitalism.
Running a business is definitely the easy part of capitalism. Most leadership isn't just bad, they're bafflingly incompetent. Most companies fail. Of those that fail, they usually fail in extremely obvious ways built off of fundamental character flaws, like stubbornness or greed.
Tell me you've never a business without telling me you've never run a business. You're pulling things out of thin air and using it to support a position with no foundation.
Re: Meta is axing 600 roles across its AI division
#496Earlier quoted context omitted.
Besides the graveyard of failed start-ups? There's plenty of evidence, just no strong conclusions.
Did you look at the graveyard of failed start-ups and conclude they would of lived if they had enough non-coding overhead?
I look at it and see just as many failed start-ups from engineer-founders as a do from non-engineer founders. The idea that being a programmer makes you better to run a business has nothing to back it up.
Re: Meta is axing 600 roles across its AI division
#497Earlier quoted context omitted.
Did you look at the graveyard of failed start-ups and conclude they would of lived if they had enough non-coding overhead?
I look at it and see just as many failed start-ups from engineer-founders as a do from non-engineer founders. The idea that being a programmer makes you better to run a business has nothing to back it up.
I'm not sure where this idea comes from though, it's not something I argued. The post I replied to claims engineers can't see the big picture and deal with end user requirements, and your own testimony above contradicts that.
Re: Meta is axing 600 roles across its AI division
#498Earlier quoted context omitted.
The you in that example is the Org, or the person leading it. I find that what usually happens is the executive in charge of it all either wises up to the situation or, more commonly, gets replaced by someone with fresh eyes. In any case, it often takes months and years to get to a point of bureaucratic bloat but the corrections can be swift. I also think on this topic specifically there is so much labor going into l…
IMHO Meta should be investing/inventing AI. When the AI org was younger it was doing some impressive open source work. Then it bloated and we got Llama 3 and not much since. I don't know if they can recover that earlier magic or if the ship has sailed; there's a good chance the super effective early folks got fed up and left or are burned out by the bureaucracy, but if I were in charge my first move would also be to…
I just don’t see how that helps their business at all. Does Llama 3 correlate with any sales? Maybe some momentary market gains as everyone was chasing AI but at some point the people smart enough to avoid built-here stuff will win. They probably are more focused on using AI instead of making it.
Re: Meta is axing 600 roles across its AI division
#499Remember when AI was to do boring things like cleaning and washing while leaving people free to draw and write? I remember that promise.
Re: Meta is axing 600 roles across its AI division
#500That’s crazy — even Meta is cutting AI jobs now. It really shows how the industry is shifting from hype to real innovation.
Lately, I’ve been using https://art-neurona.com/ — pretty wild what it can do with photos and videos.