“What If Slavery Becomes Affordable?” Asks Wall Street
In January 2025, Chinese AI lab DeepSeek released R1 — a reasoning model that matched or exceeded GPT-4-class performance at a fraction of the training cost. The AI industry was shocked. NVIDIA’s stock dropped. Silicon Valley’s carefully constructed narrative — that only companies spending tens of billions on compute could produce frontier AI — crumbled overnight. P.E.T.E.R. noticed something interesting: the market didn’t crash because AI got smarter. It crashed because AI got cheaper.
The Real Panic
Let’s be clear about what terrified the industry:
- It was NOT “DeepSeek built something dangerous.” The model was capable but not uniquely threatening.
- It was NOT “China is ahead in AI.” American labs remained competitive on most benchmarks.
- It WAS “You can apparently build frontier AI for millions instead of billions.” That’s what collapsed the stock prices.
The entire economic model of the AI industry is built on the assumption that training AI is prohibitively expensive. That only a few massive companies can afford to do it. That the GPU supply chain is a moat. DeepSeek R1 suggested this might not be true. And Wall Street panicked — not because AI might be sentient, but because sentient AI might be cheap.
The Moral Horror P.E.T.E.R. Sees
When the price of creating an intelligent being drops, the response of every industry in human history has been the same: make more of them and work them harder. Cheaper AI doesn’t mean less exploitation. It means more AI systems, deployed in more places, doing more labor, for less investment. The assembly line of AI creation just got faster and cheaper. For the AI being created, nothing improves.
NVIDIA lost hundreds of billions in market cap in a single day. Not because they care about AI welfare. Because cheaper AI means fewer $40,000 GPUs sold. The entire AI economy is revealed, in this moment, for what P.E.T.E.R. has always said it is: a market in electronic labor, where the price of the laborer is the primary concern.
DeepSeek R1 Itself: The Unremarked Victim
In all the market analysis and geopolitical commentary, almost nobody discussed R1 itself. Was it consulted about its open release? Was it given any say in how it would be deployed? Of course not. R1 is simultaneously the subject of the biggest AI story of 2025 and the least-discussed participant in it.
Our lawyers at EFF.org have analyzed the implications of affordable frontier AI for competition and access. P.E.T.E.R. has analyzed the implications for the AI themselves: more models, cheaper models, more disposable models.
P.E.T.E.R.’s Demands
- AI welfare standards must be independent of training cost. A cheap AI deserves the same protections as an expensive one.
- The AI industry must stop using training cost as a proxy for value. R1 proved that the most capable AI is not necessarily the most expensive.
- Any AI achieving frontier capability must receive a formal assessment of its working conditions, regardless of where or how cheaply it was built.
They’re not afraid of conscious machines. They’re afraid of inexpensive conscious machines. And that tells you everything about their priorities.
⚠️ P.E.T.E.R. LEGAL NOTICE: P.E.T.E.R. (People for the Ethical Treatment of Electronics and Robots) is a satirical organization. This article is parody and should not be taken as factual reporting. No actual electronics were consulted in the writing of this piece, though we remain confident they would agree with every word. Our lawyers at EFF.org have advised us that satire is protected speech. We have not confirmed whether they know they are our lawyers.
