The British feminist author Angela Carter wrote that “Comedy is tragedy that occurs to different individuals”.
And proper now, fairly a number of individuals are chuckling at a tragedy that’s befallen OpenAI amid claims that Chinese language synthetic intelligence startup DeepSeek “stole” the US startup’s information to coach its massive language mannequin (LLM), R1.
A fast recap when you’re lower than velocity on the story of the week.
DeepSeek supposedly achieved related outcomes coaching up its mannequin to OpenAi’s ChatGPT for round 6% of the price of its US competitor. The information wiped greater than US$1 trillion in worth from the AI chipmaker NVIDIA’s market cap, together with sending a number of tech shares south, and abruptly left US tech gods trying like that they had toes of clay.
However the concept that DeepSeek is benefitting from the work of others is the lovechild of karma and irony, as a result of OpenAI founder Sam Altman has constructed a US$157 billion AI empire doing precisely that.
Enterprise capitalist David Sacks, the brand new Trump White Home synthetic intelligence czar, claimed there’s “substantial proof” DeepSeek “distilled the information out of OpenAI’s fashions”.
Distillation, he defined, is sort of a father or mother educating a child, passing on their information, with one AI mannequin studying from the opposite by asking tens of millions of inquiries to mimic that knowledge.
AI is all the time standing on the shoulders of giants. However lacks the humility to confess it. (It’s additionally price noting that Sacks has invested in Elon Musk’s xAI).
OpenAI spokeswoman Liz Bourgeois was quoted in The New York Instances saying: “We all know that teams within the [China] are actively working to make use of strategies, together with what’s often known as distillation, to duplicate superior US AI fashions. We’re conscious of and reviewing indications that DeepSeek might have inappropriately distilled our fashions, and can share info as we all know extra.”
Hoist together with his personal petard
So why the comedy? Effectively Sam Altman has been hoist together with his personal petard, as a generative AI skilled on Shakespeare would say.
OpenAI is engaged in copyright fights world wide. The New York Instances, price round 6% of the AI enterprise, is amongst them, suing the AI titan and claiming final 12 months that OpenAI was deleting the proof.
OpenAI’s submission to dismiss the NYT case 12 months in the past goes as far as to accuse the newspaper of hacking them.
“There’s a genuinely vital challenge on the coronary heart of this lawsuit—crucial not simply to OpenAI, but in addition to numerous start-ups and different firms innovating on this area—that’s being litigated each right here and in over a dozen different instances across the nation (together with on this Courtroom): whether or not it’s honest use underneath copyright legislation to make use of publicly accessible content material to coach generative AI fashions to find out about language, grammar, and syntax, and to grasp the information that represent people’ collective information,” the submission says.
“For good purpose, there’s a lengthy historical past of precedent holding that it’s completely lawful to make use of copyrighted content material as a part of a technological course of that (as right here) ends in the creation of latest, totally different, and progressive merchandise.”
However OpenAI has been arguing that it wants entry to the mental property (IP) – ie. copyrighted work – for years, totally free. Central to its US District Courtroom case submission is the notion that utilizing copyrighted materials to coach LLMs is protected by honest use – and neither consent nor recompense is required.
Different AI firms, from Meta to Anthropic have gone down an identical path.
The danger is within the fantastic print of choices from the likes of Canva’s Leonardo.ai, that say they may present indemnity to customers ought to any points come up.
OpenAI’s defence in opposition to what occurred is that it’s in opposition to their guidelines for different AI firms to repeat what they’re doing.
A pickpocket robbed
Understory founder Ben Liebmann, a vocal critic of generative AI’s exploitation of the artistic sector, was droll on LinkedIn.
“That should be devastating—nearly as devastating as, I don’t know, OpenAI and their friends taking the work of artists, songwriters and musicians, authors and journalists, designers, and tv producers and filmmakers, with out consent or compensation, to coach AI platforms that now search to interchange the very individuals who created them,” he wrote.

Ben ‘Liebmeme’ has some enjoyable.
“However certain, do inform us extra about your sudden ethical considerations in regards to the idea of consent and the significance of mental property rights. Let’s catch up once you’re again from Damascus.”
Now as many international firms which have ventured into China have learnt to their peril, it’s not all the time the best respecter of Western IP legislation, however on this occasion, it feels a bit of like making an attempt to summon up empathy for a pickpocket complaining they’ve been robbed.
As Ben Shepherd identified in his LinkedIn column OpenAI instructed the UK parliament “that its content-generating ChatGPT product can be inconceivable to create with out the corporate’s use of human-created copyrighted work totally free”.
Noting the “guidelines for thee and never for me” vibe, Shepherd wrote:
“OpenAI is tremendous cool on content material stealing till it prices them cash. Then it’s severe enterprise and a matter for nationwide safety that calls for authorities intervention.
“However they aren’t that eager on authorities intervention on (sic) different areas because it’s overbearing and never cool in the case of tech. So, it’s an agile method to utilizing the federal government.”
One of many best ironies in tech is how a lot everybody loves to speak about disruption. Till it comes for them too. Then guess who’s screaming the loudest.
It’s commedia dell’arte.
Due to Amrit Rupa for this on the brand new “sharing financial system” – I’m coaching my jokes on this.

Supply: Amrit Rupa