I firmly believe it will slow down significantly. My prediction for the future is that there will be a much bigger focus on a few “base” models that will be tweaked slightly for different roles, rather than “from the ground up” retraining like we see now. The industry is already starting to move in that direction.
While I agree in principle, one thing I’d like to clarify is that TRAINING is super energy intensive, once the network is trained, it’s more or less static. Actually using the network isn’t dramatically more energy than any other indexed database lookup.
I program professionally, and I copy paste all the time. The difference is when I copy paste, its 10-20 lines of code, not a line or two— and I’m not fishing for a solution to the problem. I already have the optimal solution in my head, and I am just searching for the solution I already know. It’s just faster than typing it by hand 🤷🏻
Nope, it’s either inspecting the TTL of packets coming from your phone (unless you have a VERY custom setup, the TTL from devices other than your phone will be very different), or it’s deep packet inspection. I tried to trick t-mobile last year into giving me home internet on a phone sim, so I did a whack ton of research.
How is it scammy or gimmicky to put ads on a product you don’t pay for? I guess I’m just missing the disconnect here. This service costs Nvidia money to run.
You don’t need to exercise your stock options to access their value. It’s common practice to take loans out against their value, which allows you to access your money effectively tax free by instead paying interest against the loan. This is (again) a fairly commonplace practice used to make collecting tax difficult, and allow them to make the argument to regulators that they aren’t actually being paid that much, it’s totally just options they would never sell off. That’s why C suite has such a “burn everything to the ground, as long as our stock price goes up” mentality, because if it doesn’t, they have to start worrying about interest on their loans— because they have fairly low liquidity (percentage wise).
Base pay is a meaningless figure in tech. Total comp is all that matters. The true high rollers intentionally keep their base pay low to dodge taxes. This doesn’t mean they aren’t still getting and using this money, it’s just an intentional straw man to allow them to make this exact argument.
You can attach a fake one in software via XVFB (X Virtual Frame Buffer). It’s a little involved if you aren’t familiar with X, but it only took me an hour or so to get setup. Then you don’t need any hardware at all, and can set whatever resolution you’d like.
If that were solely true, there would be a lot more competition in the field right now. Amazon, (and to a much lesser extent the other 2 big names, GCP and Azure) are so massive not because they have a lot of power (plenty of other companies like digital ocean or OVM have plenty of scaling power too)— but because the integrations between their products are so seamless. Most of that functionality has a foundation in FOSS software that they’ve built on top of.
Regulate does not equal stop, or even really slow for that manner. There are a number of measures we can mandate that wouldn’t slow any real research, but that would curtail malicious activity, like mandating some form of detection research to go alongside models, or pushing for better watermarking technology for genuine content.
That would be a dream come true for Youtube. Think about how that would affect them financially, they would be serving significantly less load, but they would still rake in all those Youtube premium subscriptions. It would effectively be campaigning to remove the users who aren’t monetized from the platform.
The feds should claw that money back. I’m sure they intended for this money to go back into the community, and this is borderline fraud.
No, they messed up. Regardless of user count, and economic context, there is a limit to how fast you can grow a company. Going beyond that limit means that you’re diluting internal company knowledge so much that everyone just ends up doing their own thing— it’s chaos. Quality control, standards, procedures, etc go out the window. You also loose your ability to create accurate, data-backed plans with a high degree of confidence the farther you get from where you are now. You can predict the impact of a few new hires pretty easily, but hundreds, when your current team is only a couple hundred? You simply can’t forecast what holes you are creating, and challenges you will encounter with that many new people (specifically, that high of a growth percentage) in that short a time period. Growing that fast is incredibly risky, and in almost all cases, a terrible business decision. I’ve worked for SEVERAL companies that have worked this way, and it always destroys the company from the inside out.
If you want confirmation, just look at their product offering. Discord has consistently come out with features that no one has been asking for in a desperate attempt to monetize their platform, while for some reason continuing to hire like crazy (I.E., spending shit tons of money). Instead of working on their core product, and finding a way to monetize that. I can (and do) pay for all the messaging platforms I use out of principle, and I would happily pay more if their platform were more reliable. They could easily gate features in a way that generates them money, but instead of doing that, they let the core platform stagnate and add all this paid crap no one wants, of course they aren’t making money. This is a direct result of their company having a huge percentage of employees that do not fundamentally understand the product, because they hired too fast and diluted their internal knowledge.
It was never a scam, it was always a successful money laundering operation for the rich while covid had all the real galleries closed. It did exactly what it said on the tin. It’s just unfortunate that some people thought it was a real economy to begin with. Either way, good riddance.
Not sure if this is sarcasm or not, but genuinely I would be ok with this.
It discourages innovation more when the vast majority of modern cultural works, or mechanical inventions are completely off limits to all but a small ownership class. It means no one is generally free to iterate on modern ideas without risking legal liability. Ultimately, pretty much nothing in this world is original, everything is an iteration on existing works. When anything approaching a modern work is restricted, there’s far less to iterate on.
Yes, that was what I was getting at. Not having true random is one thing, I understand (and like) that implementation. Apple has been doing it since the first few iPods. But Spotify “shuffle” isn’t near even, it is exactly even, as in “if you shuffle play this playlist twice two days in a row, it will play the exact same order”. Which is why people are complaining about Spotify specifically.
The problem isn’t that their random is biased or has rules, the is that it is entirely deterministic, to the point where it will play the same exact songs, in the same exact order for days. It’s as if shuffle just activates a hidden “shuffle” playlist that only updates once a week.
No chat app needs a desktop App, they need a WEB app. Generally I’m against them, but in this case it makes sense. It makes cross platform trivial, and you would never really need to use a messaging app offline anyway, browser APIs have come a LONG way. It’s also Google’s core competency. So yes, I believe they 100% have the tools if they wanted to try.
As for integration, my point is: why would Apple bother integrating with Google’s suggestions? Google has a track record of abandoning standards and ideas at the drop of a hat. Why on earth would Apple spend time, money, and engineering talent on something that’s likely to become abandonware in 2-4 years time? That’s also assuming it’s a GOOD standard, most of the previous attempts had fatal flaws that made the product dead on arrival. If Google had something compelling, and gave us a reason to believe it would be around for more than a few years, I’m sure adoption would go through the roof, and Apple would want to integrate— Because it would now benefit them, they would be getting something out of the deal; More features, an established user base, etc.
Indexing and lookups on datasets as big as companies like Google and Amazon are running also take trillions of operations to complete, especially when you take into account the constant reindexing that needs to be done. In some cases, encoding data into a neural network is actually cheaper than storing the data itself. You can see this in practice with gaussian splatting point cloud capture, where they are training networks to guide points in the cloud at runtime, rather than storing the position of trillions of points over time.