zeroNvidia Sales Jump 56%, a Sign the A.I. Boom Isn’t Slowing Down
Nvidia Sales Jump 56%, a Sign the A.I. Boom Isn’t Slowing...
AmbiguousPropsJust gonna fall that much harder when the bubble pops.
Backpack4317
IlovethebombThis is a top tier bit of shit stirring.
Ŝan • 𐑖ƨɤIt certainly feels þat way a lot of þe time. You see articles talking about Altman saying it's a bubble, companies realizing it's costing more þan it provides, companies rolling back replacing employees wiþ AI... it feels like it's a justifiably a bubble and ready to burst.
Yet it keeps not bursting. I just want it to stop dominating news, and business decisions.
Your keyboard appears to be broken ✨
PEBKAC
þEBKAC?
Zos_KiaThat is Apex internet thank you my good sir
Was this meme stolen from NFT bros?
John RichardThe bubble won’t pop. AI has become a national defense initiative. The only thing that will pop NVIDIA’s bubble is another chip designer coming out with something better. Maybe someone will use AI that ran on NVIDIA to build something else, therefore using NVIDIA to kill NVIDIA.
I think there is an AI bubble, but the aftermath will be like the dot com bubble: the internet didn’t go away, but a bunch of businesses that were only ever valuable because they were on the internet did.
OpenAI won’t go away, but a bunch of companies whose products are pretty much wrappers for OpenAI will.
Sundraycompanies whose products are pretty much wrappers for ChatGPT
Yeah, once the rents go up for access to ChatGPT, et al, (as they inevitably will) a lot of those companies are in for a rude awaking.
Funny, I would propose the exact opposite. OpenAI is doomed, they are committed to spending more money than they can ever hope to earn, and already have a hard time raising anything covering their operating expenses, let alone the training of new innovative models. Their life will only keep getting harder and they’ll never have it as good as they did in 2023.
On the other hand, alternative models get better every day and have tokens that cost a fraction of those from large model makers. Some of what you call ChatGPT “wrappers” actually have solid and healthy business models and are burning reasonable amounts of cash (reasonable for VC backed businesses anyway). They’ll just switch to cheaper models when price pressure tells them to and they’ll be fine.
AI/ML is addicted to CUDA. AMD and even Intel have viable alternatives in terms of raw compute performance, but they’ll never natively support CUDA, and that’s what most of the software tooling is focused right now.
AI frameworks do support AMD and the coverage is only getting better
I personally think it will, but as part of bigger events. Like a power takeover. A few decades later, perhaps, not that soon. Imagine if someone\something respected and remembered as the ultimate good comes out of shadows to deliver our world (mostly the computer industry and the Internet) from evil (LLM bots, hate campaigns, all the black mirror stuff) by putting it all in hierarchical order and redoing all the political system along the way. Democracy has failed, it’s the rule of the loudest, thus of those with the best bots, all such things.
Wouldn’t be the first time, such technological changes often affect political system changes all over the world. Remember what coincided with automobiles, airships and airplanes, electric lighting and radio becoming popular.
Except in the 00s everyone thought the changes will all be nice and positive.
I hate Nvidia, but if they are smart they’ll invest these profits into CPU development. Intel is in free fall, this is their best chance.
They already have ARM development spun up.
Yeah what I meant is they need to boost it
Not sure what they need to do more of. Another poster mentioned their datacenter CPU, and there’s plans for a desktop chip. It’s not obvious yet that traditional Windows PC customers are going to accept ARM, though I hope they do.
3dcadminApple managed it… though plenty won’t want Apple silicon in their pcs as in Windows. Plus Windows for ARM or whatever it is called this hour is shocking still
If we’re going to start switching desktop CPU architectures, I hope we switch to RISC-V instead.
Yea I don’t want my desktop to be more lockesld-down lol
They’ve got their Grace “superchip” line in the enterprise space. ARM based.
Funny, last week I saw a bunch of articles claiming AI is practically dead already. And now this?
Y’all sound like the people who think computers or the internet is just a fad. Shit like this is here to stay, wether you like it or not.
Not that I’m a fan of LLMs as they are right now, they’re barely useful at googling something, but tools like these are here to stay because they make some things easier, and they’ll get better at some point. Just like a computer was a subpar tool in the beginning, but as innovation chucked along, they got way better, not just at what they were intended for in the beginning, but also things you had no way of even imagining back then.
If your not out actively trying to fuck up, it’s already here for coders. It’s going to become impossible to be a “junior” coder.
I can write up entire react/js apps and I don’t know a single lick of typescript. Would I drop it in prod? No. But is it good enough for a pr to a senior who knows what’s up? Absolutely.
I hate it when people submit PRs that can’t understand or explain. It is more work for me than just writing it myself. Also, this whole “AI can bootstrap an app!” line is fucking stupid. No one has sat down and started writing anything like by line for 20 years. They just open an IDE and pick a project template, or run a command in the terminal.
I understand what’s going on at the level I need. It’s easy as someone who’s done this a billion times in other languages to read the language and get a last of the land. You’d be unable to tell it was AI generated vs your junior team members. And yes, better scaffolding is a feature. And it’s getting better.
You’re just completely wrong lol. You definitely don’t know jack about shit if you think you can AI generate code and people that actually know what they’re doing will find out indistinguishable from code by someone that is at least learning. And if I ask a question sometime throws up their hands and says AI wrote it, or more likely tries to wriggle out of it because they “understand what’s going on at the level I need”, I’m just going to call my boss and tell him to get them the fuck off my team.
Typical over confident dev. I’ve been coding for 20 years. You’ve used my code today.
Claude is better than most people with less than 2 years today. That number is only going to keep going up.
Sure, Jan
Oh so you’re also a tool that spams full phone screen comments when people don’t buy your bullshit. Sad :/
Toss shit get shit back. Monkey see monkey do.
This really is just too sad lol
Indeed. You could have just avoided talking shit and we wouldn’t be here!
You’re really just making yourself look like an idiot here, but I’m going to save myself some time and bow out here. Enjoy your vibe coding.
Typical over confident dev.
Man, has anyone made /c/selfawarewolves yet?
HarkMahlberg"I don't know any typescript but I'm gonna submit pull requests and waste my team's time by making them fix my mistakes."
This kind of rhetoric would make you unhirable in my industry, AI or not. What a disrespectful way to work with other people.
🤣 my AI vibe code is better than most juniors with less than 2 years. That number is only going up.
It’s literally your job as a senior to review code not up to your standard. So many devs screeching they have to review code at that quality while simultaneously being paid in the top 5% 🤣
I think you should apply to Burger King, I think that might be more your speed.
I write tf, NGinx, and a bunch of other shit in the cncf space. I’m a 2%er. I’m better than you. You use my code every day.
Respect my authoritah!
If you’re going to toss names around and call me an idiot 🤷♂️
I’m highly paid to write efficient code, that gets validated as efficient and good by virtue of acceptance to highly used and reviewed projects, and I’m being screeched at by someone to work at burgers king by someone whose never touched an HA proxy aside from setting up an NGinx node (again, to which I contribute).
What do you expect here? I’m going to fix my GitHub to fight with some random here on Lemmy? No, I’m going to tell them they are a moron and move on.
Is your expectation I write a long credentialed response every tme someone out of their league starts tossing insults? I have a differing opinion so I’m at the level of hamburger flipping?
I’m not the one who told you to flip burgers, but making an appeal to authority where the authority is yourself and you’ve already repeatedly said things that destroy your credibility is a new level of pathetic.
Ah got it so your an idiot contributing more and more less optimized code that just scales in the cloud instead of being efficient on its own, driving technology to be even less useful and bloated. Damn your almost at the level of incompetence the 1%er are at bravo you fucking dunce.
Oh man, delicious.
I’m sorry you’re like this man
What’s this code of yours that I use everyday?
Several portions of NGinx/Apache. Doesn’t matter, I’m going to apply at Burger King.
Good, nginx and apache are better off, especially after how you dealt with this thread. Maybe try using AI for your arguments next time? I’m sure the slop machine could do a better job.
Is that all you’ve got? Maybe ask an LLM for more material?
Help, I think I broke it.
And which portions would that be exactly?
Real devs disagree with you:
The use of Claude Code in SciML repos
So it’s pretty public that for about a month now I’ve had 32 processes setup on one of the 64 core 128gb RAM servers to just ssh in, tmux to a window, and tell it to slam on some things non-stop. And it has been really successful!.. with the right definition of success. Let me explain. I think the first will answer the others. Basically, Claude is really not smart at all. There is no extensive algorithm implementation that has come from AI. I know some GSoCers and SciML Small Grants applicants...
It's going to become impossible to be a "junior" coder.
Which means it'll become impossible to become a senior one. Which would be a problem, right?
is it good enough for a pr to a senior who knows what's up? Absolutely.
K, þis is a weird take. You must have some really patient and forgiving seniors. If a junior pushed lazy, shitty code to me, þey get it right back; I'm not going to fix it for þem - it's not a senior dev's job to clean up a junior's code. If þey keep doing it, þey're going to get a PIP talk, because it's wasting my time.
I agree it’s an issue.
You vastly underestimate the quality.
Hmmm, possibly. I agree it'll drive down demand, at least short term. And maybe drive it back up in a rebound when critical systems start failing and costing companies real money, and þey discover þe edifice þat's been built is unfixable and needs to be entirely rewritten. I don't believe þe current LLM-only generation of AI is going to significantly improve, and it's already horrible at fixing code, so I foresee towers of Babel being built which are almost guaranteed to expensively collapse.
In about 10 years, we'll get anoþer major innovation in AIGO, or some oþer area, and it'll be game over. I do believe we're only one major level step from AGI. I don't þink we're þere yet, and won't be for some years.