There is zero proof of distillation. Minimax 2.7 development was surrounded by moderate use of Claude. M3 is their latest generation, and pretty solid, but its performance cannot be attributed solely (or even 5%) to distillation, and that is only lab that has been accused of significant API use. These claims are all 3-4 months old by now, and Anthropic blocked China access after publishing the accusations. Repeated BS is BS from losers trying to lobby for support.
Also Cursor (recently bought by spacex for $60B) “stole” not only the entirety of kimi 2.5, but all of their users data to train its composer models ontop of kimi 2.5.
article ignores the circular financing bubble. It layers on top all of the hidden costs in servicing their fake revenue and fake stock valuations.
A bigger factor in AI fraud economy, is the circular fake assets and fake revenue. $1T valuations in top 2 LLM providers implies $50B-$100B eventual recurring profits with little to no reinvestment. $1.75T SpaceX valuation was based entirely on fraud of at least 5x more expensive space data centers than earth and supposed 90% of value of company is based on Grok enterprise TAM. IPO was boosted with fake Anthropic paper lease where they were renting gpus for 5x the cost, and over 2x the list price of on demand rates from Nebius/coreweave, but with instant cancel clauses, and “first 2 months free” scam. 11th hour similar Google deal (who owns 6%+ of spacex) further amplified the SpaceX fraud IPO.
The main circular fraud though, is most of the companies (Nvidia not mentioned) in article put an inflated asset on their balance sheet, while trading revenue credits for those investments. Article’s “footnote debt” is largely to serve those revenue credits.
There is no evidence of “theft”/distillation by latest models. Anthropic’s original allegations were weak against deepseek. When someone publishes vibe code to github, it’s not theft from the LLM to train on it. Repeating loser whinning often enough to make it US dogma, doesn’t make it truth. China is a short fuse away from sinking US fleet blockading Iran.
I don’t get the general criticism. It will all depend on details. If you can “rent” a $1000 device for $500 in payments over 2 years, and have the option to pay $500 at end of 2 years, or pay $250 over the next 2 years, then that is a better deal for you, including the delayed payments, and your optionality. If rampocalypse is over in 2 years, you can get a better device. How far away from the above formula implementation is matters though.
An advantage with buying a “good phone” is that Apple needs to keep their trade in incentives high to get you to upgrade later. Lease provides a clear upfront proposition instead of hoping for future Apple generosity.
The AI hype/scarcity frenzy (bubble) was based on Meta and xAI hoarding GPUs for themselves instead of reselling compute, making it abundant instead of scarce. All of the other datacenters were hoping to have them as customers instead of competitors.
2 lies in article:
byline “proves compute is scarce” is opposite. This is not a real datacenter rental deal. It is a “can quit anytime” show deal. Token consumption is down 20% since march peak.
“Compute costs are rising exponentially”. Blatant lie, in terms of compute pricing $/gpu hour. Rates are near their accounting floor (though not at rock bottom) for H series, and the faster newer cards are just 2x the $/hour instead of the 8x saturated performance difference.
You see… the previous corrupt process was only a problem not because doors were falling off planes mid air, but because that happened recently enough for you to remember it tied to the corrupt process. If Boeing isn’t allowed to sell cheaply made death traps, then China and Iran win.
Clitoris is that girl who get’s raped on a pinball machine in Silence of the Lambs.
This seemed more alarming news: AAI is tracking IPs/proxies/vpn, timezones on requests.
https://tech.yahoo.com/ai/claude/articles/anthropic-caught-secretly-spying-users-204222066.html
While it is explained as “only to accuse China of evil”, its tracking everyone, and empowers US government crackdowns on anyone either AAI or US doesn’t like.
For last 5 quarters, Nvidia has sold much more GPUs than get installed. On 3 year depreciation schedule, 8xh200s are $1.50/hour per card just sitting there. B200s are $2/hr. (Nvidia changes network cable each generation). Very expensive to do what you are suggesting is happening, but only other explanation is smuggling into China, or secret sovereign/military purchases.
It’s also possible that they are lent to tier 2/3 datacenters in revenue share deals.
The fundamental lie behind the AI fraud is that compute is scarce relative to demand. OpenAI, Grok, Meta did overinvest in hoarding GPUs far ahead of their usage. Last 2 are now competing on hourly GPU market, and AI tokens/revenue has fallen 20% since spring peak, which is a bubble pop compared to 10x/year perpetual growth expectations behind the hoarding. While hourly gpu rental market is stable rather than declining, it is stable at very low prices, mostly for accounting reasons of not actually losing money intentionally. Deployed GPUs are abundant proven by accounting floor based pricing. A b300 can deliver 8x the tokens of an h200 but only rents for 2x more.
most of the fuel weight required is to lift the rest of the fuel. Fuel costs is about $1m for full load. Rest of cost is huge staff, maintenance, and capital cost of rocket.
The $200/kg launch price target is based on 150 ton capacity. That’s a $30m launch costs target. Volume/foldability matters the most because that is the actual constraint that limits datacenter launch to a single NVL72 size.
It will never be an economic thing. Only unpluggable skynet military thing. The weight is not an issue. though. It’s volume.
Incredibly bad, too lenghty, mostly irrelevant criticisms of the fraud of space datacenters, followed up by the Skynet military justification of being unable to unplug skynet.
Space datacenters from SpaceX are a fraud because they have a 5 year lifecycle with deorbiting of entire unit. The costs compared to 30 year lifecycle of terrestrial solar/battery powered datacenter energy is thus 6x higher (costs of shell/shield, solar, radiators is about the same but 6 replacements). Terrestrial building costs are $20/watt. SpaceX ambitions are to get $30m/launch costs. To be only 2x the terrestrial costs, launch costs need to be $1m (just the fuel costs) with deorbit being to fly off into space instead of a salvage trip.
At 12x the costs, the competitive GPU rental hurdle has to be 12x more expensive than earth. Only military skynet applications would pay for this, and specifically, only permit mechahitler to decide if skynet is doing a good job.
Gulf Cooperation Council is essentially west side of Persian Gulf. It is very oil rich region, includes 2-3 very rich, countries with city infrastructure better developed than any US city. Dubai is epitomy of this with global financial/transport/tourism center, tallest sky scraper in world. War has caused a 30%-50% property value drop. Iran has been striking energy and US bases in these countries.
US wealth is far more dependent on GCC investment than Israel control. This war has ended, because GCC were threatening US. Terms includes not rebuilding destroyed US bases, and leaving the ones they still occupy.
Headline is innacuralte. Customers who purchase a $200/month subscription, can save up to $13800 on token costs compared to “a la carte” token pricing. Open AI does lose over $1 per $ in revenue, but this is more a function of them having too much compute, and very high training costs, rather than a loss per gpu hour on tokens served, it should stilll be a loss for OpenAI:
Open source model pricing per gpu hour, ranges from 80% margins at high batch saturation per user (low tps per user) to 50% at concurency of 8, to loss at fewer user requests per gpu. https://inferencex.semianalysis.com/compare/deepseek-r1-b300-vs-h200
OpenAI has very high prices, and too much compute. They would likely lose $6000 per user who maxes out their $200 plan, and certainly over $1000 per user.
It is burying old news in headline. The consequences/pitfalls of their “strategic shift” are fairly new. The layoffs were explicitly justified for pivot to datacenters that OpenAI will “surely” be able to rent. The new extra problems in that strategic shift just makes them look worse for going all in on the bubble.