Indian Researcher at Anthropic Quits Job to Pursue Poetry

AI startup Anthropic hogged the headlines ten days ago, after $300 billion in market value vanished in about a single trading day after it launched its chatbot Claude Cowork, a set of 11 open-source plugins, that help AI agents complete tasks autonomously.

The next day Anthropic launched C compiler Claude Opus 4.6 model, which industry experts said was not a normal task and alerted coders to find alternative jobs.

Anthropic is in the limelight again with the sudden exit of one of its young Indian origin researcher from the company. Mrinank Sharma, an AI safety researcher at Anthropic, recently announced that he had quit the company.

In his cryptic post on X, the young researcher did not reveal the true reason behind quitting his job but hinted at moral responsibility and poetry. He indicated that he was disillusioned with technology as he points out to a gap between what companies claim in public and what they practice at the workplace, raising doubts about AI’s transparency.

Having completed his doctoral studies in machine learning from University of Oxford and a master’s degree in ML from University of Cambridge, he has a firm standing in the tech world. Then what drove him to quit the lucrative job?

In his message, Mrinank conveys that the world in facing several crises at once. It is not from AI alone but from other interconnected problems. He says that in the upheaval, someone needs to make sense of the world and that he would try to do that by writing poetry.

Mrinank says instead of teaching AI to be more transparent, he feels called to writing which would allow him to place poetic truth alongside scientific truth. In the post, he expressed readiness to explore a poetry degree and devote his time and energy towards courageous speech.

The entire post reflects Mrinank’s disappointment over not being placed in a job that would give him the inner satisfaction.