motbus3
It seems to me that nvidia bubble is still to burst. There is a trust me bro circle where Nvidia somehow sold GPU to itself so it can host someone else's models. Maybe three years ago or a little more they announced they plan to move away from consumer end and focus on cloud. Back then it did not make much sense because everybody though "how the f Nvidia will manage clouds like AWS?" And their business model seems to me that it is a shared risk model where Nvidia don't "own" the cloud but pay for its GPU to have that market share. I think it would make much sense if the thin layer those, how did the greedy musk said, neocloud/cloudfronts or whatever, were not so brittle due to the lack of market demand.
I have the impression they had a gold strategy for them but for some reason they rushed it too much and now they broke the market before they could profit from it.
I can imagine that if they didn't hoard their own resources to create scarcity artificiality, they would possibly have a stronger position in the market. I can only think they did this because (A) they know the current government will bail them out and (B) someone was worried that China could hoard those GPUs first.
But as I said, I don't understand much of the topic, but it seems to me they are try to get as much money as soon as possible and that looks scam-ish to me. They all put investment rounds too close to each other, started creating concerning stories and "rogue" AIs at the same time. Maybe I'm being too paranoid, but so far those guys have been pulling some hat tricks and I think I know who will pay the bill
This hits people much more than only getting less money.
But this is just an example. Unfortunately, Argentinean people are now brain-dead and they are quite lenient to corruption and puppet governments anywhere. I find it funny for them to complain about Malvines/Falklands while their president is American.
I think we might see more soon. I think the fact the are hiding the harness and the "real" model is a signal they are not confident they will keep improving. openai reported 95%+ on some benchmarks with an unpublished harness but when running with public harness and public model it is not a 30%+ improvement it is a 6% if that much...
I really wonder why they need to do that
The only thing community driven I used nowadays is for cooking because LLM recipes are always the most generic thing. (Though I use LLMs for converting metrics or question if something goes wrong)
Software engineer is far from be solved
Your comment about academics is not based. If you take the time to read any of the papers they published you will see most if not all are pure garbage nonsense.
Read Terence Tao about ai, or anything about the "supposedly" improvement in the FFT. Their claims are all nothing burgers.
There is nothing that guarantees now that they will not steal your ideas and publish themselves.
I dont know why it is taking so long for investors to catch up with this suspicious behaviour
I think now LLMs might play a part of those dismissals but also I think those companies are embracing for thin cows season.
The record profits and etc are the last few for a while I am guessing. My guess they might have up to a year of positive numbers as consequence of deals and layoffs up to this moment.
I think at some point it will be clearer what are the real productivity gains of LLMs. My opinion is that there is a gain but we are already on the phase where there is a deceleration on the improvements and the costs have not yet caught up. The reason for this is the frequent change in the reporting of results. Now they hide the models, they hide the harnesses, they hide the results. They would not be doing that if they were 100% confident it was working perfectly. Also they have been putting too much effort on blocking other initiatives by hiring important researchers with no clear proposition on what they will be doing. One might argue that when you hire Karpathy you never know what he will might be doing anyway, but I digress. We also see some of the huge investment funds starting to switch strategies.
Unless there is no breakthrough, we might see LLMs results in a couple of years. The company I work for and others from friends I know have all started restrictions on costs. Some even started hiring people back.
My prediction is that the low AI cost per token era will end very soon when investors realise that the market is broken and they want their money back for leverage themselves or other businesses.
I even think this is why Elon Musk is merging all his companies. He might making himself artificially richer, but I think he is focusing his efforts to smart weapons and robots.
I think there will no way back to the inflation of people as in 2020/2021 but probably we are going back to early 2010s number.
I understand that US is migrating to a no friends policy, but I don't think its internal market is nearly ready to support it if clients start to go away.
Anecdotally, I talked to people in a recent event and the common topic was "what we are doing to avoid a blackout of US tech." People on big companies have reserved hardware and alternative software and are scaling down Amazon and Microsoft solutions at a fast pace. People on smaller companies are doing similar strategies but relying on a multivendor strategy.
I think this was inconceivable 3 years ago. Every company in Europe and Asian would gladly pay for amazon for cloud services and now they are using AI to migrate out the lock-ins.
We are seeing American LLM companies more and more relying on "secretive" methods to increase scores and this can only mean that they are not evolving fast enough. The gaps between the reported benchmarks and measured ones are increasing. It might be "safety" concerns or whatever, but I don't buy it.
Like Musk, Altman and Amodei have been overpromising every time they need more money and each time they need more money and now they internet is poisoned with their own content and they can't even trust any sources anymore.
I dont know when, but i think if you keep menacing of taking technology of people hands for any political reason will be a reason for people to not use it or to rely in a more reasonable party.
The whois marks everything to private
Then you can follow things and you can see they receive 1.10m in donations https://register-of-charities.charitycommission.gov.uk/en/ch...
And there you can see this guy as the charity chairman
https://www.theguardian.com/business/2022/may/14/nigel-hugil...
And so he is a house developer? Already sounds a bit of conflict of interest. But then let's check who is the company listed
And Gatsby Is listed as a main source of income for both company and charity parts
Which is a charity by the Sainsbury family https://register-of-charities.charitycommission.gov.uk/en/ab...
I will dig this when I have time but this centreforcities is not an unbiased source of information
I would say that we probably would want someone around 40s or older for that position applied (or I guess so because we never get to ask those things)
I was in the step before the interview with the CEO of the company, and we approved 10-20% of the candidates. First steps were approved at maybe the same ratio, and the CEO decided not to hire anyone in the UK after trying for a long time.
We had only maybe 6 candidates for the last phase, and out of those, there were only 2 candidates we really thought could be approved.
I will say thought, we were committing a sin that now many companies are doing. We were expecting too much from the same person. I am sure that one person would be able to hold what that position required but I don't think the professional with that particular set of skills is easily available in the market. I guess money wise the company did not want to hire more than one person, but if it was me, I would have hired a senior instead of an specialist and a junior.
But what I am definitely afraid of is that they will want to charge games per hour anyway. The talks that consoles will cost 2000 USD is part of it. They don't want you to own things anymore.
You can't buy a house anymore, because they rather buy and keep it to create shortages and make you pay an absurd rent.
We should be talking more about how taxes should fit into this world where they are positioning themselves to not allow people to have things, not even the basic.
For examples, in the housing thing, it should not only have higher taxes for keeping properties empty but also it should be exponential otherwise it will become a cost to people. Making it exponential make it unviable. Also, investment funds should be much more regulated and should require approval for every transaction above a threshold (which imo should be small to close loopholes). This will slow things down enough.
The demand will still exist but companies will fill the gap. We should not be relying on one single investment fund to control your health, housing and employment.
We also should be discussing on closing all the gaps that ultra rich are using to not contribute back to the system they explore. They became leeches/parasites and not putting back the amount they should is causing them to be able to break the system more and more.
The only Pacific way right now is to boy;cott these big fellas. They got to comfortable saying that we are trash and the blame of everything is on us while doing everything they can to not even pay basic rights to employees.
I also think that if you know a trillionaire employee with a white collar, they also should get bit more pressure.
I know someone working in a company which recently caused hundreds of "collateral" deaths, and for him and everyone in the company, it is not their fault they exploded innocent people. We need to put these folks back to reality, because they are the same type of trash as we are, and they don't know it.
Now I guess it is obvious why they have been building so many bunkers lately. They were definitely conspiring.
(Btw, I have a strong feeling that sxai+t might become an investment fund with several companies, but that's another conversation)