OpenAI has ChatGPT, Anthropic has Claude, Google has Gemini. But what does Amazon have?
Not a whole lot, it turns out. Unlike its competitors in the tech megacorp scene, Amazon’s standout AI model — organized under the name “Nova” — is about the farthest thing from a household name. And that’s not likely to change anytime soon, because the ecommerce giant is now gutting its in-house AI labs, Bloomberg reports.
Per the outlet, Amazon is scaling back the ambitions of most of its Nova AI models and reorganizing its AI development teams as part of an effort to narrow its focus on “highest priority” goals, after its power-hungry large language models turned out to be a bust.
Accordingly, the company has put most of its AI models on ice, Amazon insiders told Bloomberg. This includes its text-based Nova models, as well as video- and image-generating models, all of which have been placed in a state of “keep the lights on,” meaning they’re technically still supported, but only receive the bare minimum resources needed to sustain them.
Instead, labor and computing power are being diverted toward a singular “frontier-model effort” led by Pieter Abbeel, director of the Berkeley Robot Learning Lab whose robotics company Covariant was hoovered up by Amazon in 2024. At the time Abbeel joined the tech giant, Covariant’s founders were working on AI models for robots, which lines up pretty well with Amazon’s push to automate everything it possibly can.
News of the shift in Amazon’s strategic focus comes a week after it closed one of its key AI offices in San Francisco, an 80-person site specializing in research on artificial general intelligence, the supposed next-level of AI development where the tech obtains human-level intelligence and reasoning skills.
The moral of the story is to sell the shovels
And when everybody is rushing to sell shovels, you gotta sell wood and iron.
And when everyone is selling wood and iron, you burn the forest down.
The shovel sellers (i.e. Nvidia) started lending money to the gold diggers to buy shovels.
The boom will end and the bubble will burst. Hopefully at least one of those companies goes under. Hopes it’s Google and Meta.
Oracle is the one that has “bet the farm” so to speak.
Also led by the biggest ahole of the three
It’s a shame Oracle’s downfall couldn’t have happened a decade ago. Ellison has branched out and bought up a bunch of media companies in recent years and is setting up his (somehow even worse) son as heir to his empire, so even if Oracle dies we’re still stuck with the fuckers for the foreseeable future.
He’s been detested for decades now, I’m sure there’s a cohort of greyhair tech workers that’ve been waiting a long time for his downfall. Hopefully he’s leveraged so badly that when the bill comes due, he’ll have to give that shit up and leave us alone.
Google is printing so much stupid money that they can still (almost) pay the AI hype from their earnings. Google will (sadly in my opinion) not go under.
Yea, they’ll most likely survive this, but I am hoping they get broken up for being monopolistic (it was a tragedy that Judge Mehta did follow through). Buts that’s a different topic. Or that they go the way of other giants like IBM and Microsoft. Not gone, but not what they once were. Though I do want Google and Meta gone or at least highly regulated, but that’s also another topic.
They won’t fail, but they will change eventually. Capitalism demands it.
Youtube ads are now unwatchable, they’re trying to wedge people off adblockers but they’ll end up leaving chrome.
Google AI search results cost a LOT more than their classic search results. That’s not a gain in profit.
Everything they’re doing is WAY more expensive and it’s just in the name of trying to prove to investors that they’re in a growth market. AI will stop being allowed to be considered a growth market.
If Youtube isn’t growing, and Gmail isn’t growing and Search isn’t growing, their valuation will plummet.
All of a sudden you’ll see a LOT of firings and a LOT of cost reduction in places that don’t make money.
It’s gonna be OpenAI and Oracle. Google and Meta (and Microslop) are too diversified to go down, even if they will take massive hits.
Neither of those are likely. OpenAI, Anthropic, and Oracle are all likely to go under, to then be bought out. Facebook will take a hit in stock price (maybe) but they have profitable divisions and Zucc structured the stock so he can’t be voted out. Amazon, Google, and Microsoft will likely see some executive shakeups, possibly to the point of the CEOs being ousted, but they all also have profitable business lines so the survival of the companies themselves is not really at risk.
Guarantee that Meta goes under. The only profitable ideas they’ve implemented were either stolen (Facebook) or bought from someone else (Instagram). Meanwhile Zuck loses hundreds of millions of dollars on dumb shit like the Metaverse
The shovel industry is booming!!!
Well according to your own link the bubble isn’t going anywhere
Not long as these two companies continue to make bank from this bullshit.
Nvidia is doing some real shady stuff, like selling GPUs and then renting them back from the companies they sold them to, which inflates their own value. That money doesn’t actually exist, though.
What makes you think that?
First off, the bubble can and will (likely) explode as soon as the rest can no longer buy Nvidia and AMD chips. If anything, what you point out only indicates that these 2 companies may not completely implode when the bubble pops
Second, a lot of the money being passed around circularly comes from Nvidia… they are literally giving money to their clients so they can turn around and buy more chips from them
You seem to confuse market value with profit. These numbers are pretty useless for any gauge of what is really happening.
Step 1: Spend tons of money on AI Step 2: Fire 10-20% of total tech workforce Step 3: ? Step 4: Profit (post positive earnings) Step 5: Fire even more employees because your AI solution is failing anyway
Amazon both fired a ton and had a ton of senior people leave, they burned the candle from both ends directly and indirectly and the effects are starting to amplify.
lol
Lmao, even
WAIT, you mean not every single LLM company is going to win the LLM race? That in the end it’ll be somewhere between 0 and 3 companies who make big money with a handful of niche players? Who would have thought?
Amazon is playing the medium game.
Pivoting to build up its drone amd space force before attacking.
Kind of funny in a way.
The ultra wealthy hate normal people that don’t have a mental illness regarding hoarding. They sabotage education so that the populace is stupid and votes for them. Then, they create AI with the hopes of learning from the stupid populace, and it fails because the AI is synthesizing stupidity.
Maybe the stupid MAGAs will be their own downfall
Oh noooooo
Anyway
yet we pay for their failures via our tax dollars… Billionares are the welfare queens.
I’d rather have Democratic Socialism than MAGA Socialism.
How are our tax dollars going to Netflix?
Edit: oh, it’s about Amazon
Amazon has had a huge brain drain going on for the past several years now. They’ve lost a huge portion of senior engineers to competitors and replaced them with people too junior to step into the hole that was left. Amazon is faltering.
Back in the 90s, I was working in the record business, on the sales & marketing side.
All these chains of massive big box record/video/book/electronics/computer stores started popping up across America - Best Buy, Circuit City, Media Play, Virgin, Borders, Barnes & Noble, etc. - as well as all the established chains like Recordtown, Camelot, etc. adding big box stores. Every city had at least one of EACH of these chains, if not 2 or 3. Every city, even smaller or mid-size cities, soon had 10 or more Big Box stores spread around, each with massive displays of inventory.
That meant every city had far more inventory than any market could possibly bear. I was in the Classical/Jazz division, and knew that even a “hit” classical record would only sell maybe 50 copies in a city, but now there were 1000 copies in all the stores in that city, and MOST of them would never sell, and would eventually get returned. But I got paid commission on sales, and my commission wasn’t clawed back if it got returned, so I did my job, and sold all that overstock inventory that the Big Boxes demanded, collected my commission, and bought a house.
I knew ALL the major players at the Big Box chains, and they were aware that the situation was unsustainable, but eventually they would run all the others out of business, primarily through predatory loss leader sale pricing, and be the last man standing. That was the strategy that they ALL had, without exception, and it worked. Before long, they all started to close stores.
Today, the only one left is Best Buy, and they don’t even carry music any more, after running every other music store out of business.
The point is, that same Last Man Standing scenario seems to be the strategy here, too. Lots of money will be wasted by companies who blink, and abandon their commitment, or are run out of business because the market simply can’t bear that much competition for limited interest in the product. At least people WANTED music, people are literally HOSTILE toward Data Center products.
And even the losers will end up with datacenter sized space all over kingdom come with access to massive amounts of electricity and water to begin the next phase of evil.
Most of them will end up abandoned, like the malls of the 90s, and will be sets for shooting zombie movies, homeless encampments, cult communities run by shadowy warlords, wild animal colonies, etc.
Amazon already won their version of that contest. I can’t imagine how the people running it think, but maybe they wouldn’t want to play that game again when they can just wait and buy up all the losers at a small fraction of the current value.
I feel for the developers who will inevitably lose their jobs. Otherwise, HA HA!
I’m actually a bit disappointed that Amazon isn’t going to crash and burn like all these other companies when the bubble bursts :(
True, but Amazon abandoning their commitment will influence others who are on the fence, and they will quit, too, thus hastening the inevitable crash.
Think of this as the beginning of the end. Amazon still counts as part of it, and they will have lost BILLIONS in the process.
So turn that emoji frown upside down! Amazon took a serious and painful spanking!
Don’t worry, they still have ways to suffer:
During the call, Jassy told investors that Amazon now expects to spend $220 billion in capital expenditures in 2026, up from its prior estimate of $200 billion, due to higher memory costs. Even at the elevated level, however, Jassy said Amazon still won’t “have enough capacity to meet all the demand we have in 2026. And I believe this dynamic will also be true in 2027 too.”
They go into debt to build the compute infrastructure for others. And then that won’t deliver any significant amount of rent, once OpenAi and Anthropic run out of money.
I really need the AI bubble to wait two months to pop so that I can get my buy-a-house seed money out of the market.
Umm, bud, get that money out now? It’s commonly advised to not bet anything you need, or use it as a savings account. Especially not now, before it was barely somewhat possible when things were stable, but not now. And with your down payment? Pretty serious stuff to lose just because some jabroni sends a tweet or didn’t bribe the administration enough. Things can go right, yeah, but think about how wrong things can go, and how fast, and how much more likely that is than making… What? 1k or something, at most? Naah, naah. You can just negotiate a better price for the house to make up for the difference.
Umm, bud, get that money out now?
Can’t. It’s a company stock plan where you can choose to buy it and have payroll deductions over a year, and at the end if the price has stayed steady or increased they give it to you at 20% under the price it was a year earlier. Since my company stock is 4x what it was a year ago that became some real money, but I think it may be tied to the bubble so I’m hoping for an october pop.
That person is right, but I think you’ll be fine. It’s only gonna crash hard after AI companies finish their IPOs and get some nice bag holders. Until then they’ll do anything to prevent it. Doesn’t mean they’ll succeed though !
Many LLM labs are shockingly dysfunctional.
Amazon and the useless Nova series is a great example, but Microsoft has its own too. AMD as well.
Europe has a ton of them, labs barely replicating obsolete Llama2 training routines in 2026, but somehow getting tons of taxpayer dollars? Saudi labs doing basically nothing.
I mean… I could follow an off-the-shelf Nvidia Nemotron recipe and do better. Or more sanely, just start with that as a base. And that’s sad; apparently these highly paid labs don’t even follow the space enough to know to do that? What on Earth are they even doing internally?
Have you seen how these tech companies treat their workers? Too many in the field are extremely burnt out to point that they can’t function anywhere near the level required for the technological breakthroughs that the execs are demanding. They want to hire entry to mid-level data scientists for the low pay and then expect them to output the quality of work of a senior data scientist with 10+ years of experience. They’re being set up to fail.
That’s not even “breakthroughs,” though.
That’s just “follow instructions for what’s already been done, and meet par.” Anyone with some Python experience and instruction following ability could do it. In LLM land, what some research houses are doing would be like releasing some new white box computer with chips from 15 years ago.
It makes no sense. It’s way too late to be interesting.
But I guess it’s getting funded because the funders can’t really tell.
Insert sickos.jpg here :)
