The Head of AI Safety just quit to become a POET …and his parting words are a message to humanity…

So yesterday, Mrinank Sharma, the clever geezer leading the ‘Safeguards Research Team’ at Anthropic (Claude), RESIGNED.

For those who don’t know, Anthropic was touted as the “safe” alternative to OpenAI (ChatGPT).

When the guy in charge of the brakes leaves the building, people notice.

…But it’s how he left that has everyone talking.

He left with a message about “The “Invisible Thread” of Integrity”..

Sharma didn’t just leave a standard corporate “thanks for the opportunity” note.

He posted a deeply philosophical resignation, claiming the world is in “peril” from interconnected crises.

He argued that technical safety (code and math) isn’t enough anymore – we need “courageous speech” and “poetic truth.”

He ended his announcement with a poem by William Stafford called: “The Way It Is.”

The core message in it is that there is a thread you follow that goes among things that change.

But ‘it’ doesn’t change.

By ‘it’, we think he is referring to INTEGRITY.

All the online gossip today is asking: what is this actually signaling??

Is this a bloke who’s just sick of all the toxic, corporate office bullshit, and thrown his job in the ‘f*ck it bucket’???…

OR…. is it a warning from somebody who knows what’s coming, and who’s humanness is pleading for all other humans to listen…??

‘As the world around us changes, hold on to the thread of integrity and human soul’, is what I’m feeling.

With Anthropic now valued at $350 BILLION, the pressure to scale and profit is massive.

Sharma’s letter suggests (not explicitly said), that “values” are becoming harder to maintain under that pressure.

He’s reportedly moving back to the UK to get a degree in Poetry following thos huge departure..

His logic?

Our technological power is growing faster than our human wisdom.

He seems to believe that the answers to AI risk aren’t in more code, but in understanding what it means to be human.

Even more interestingly; this follows a long line of “safety” researchers leaving major AI labs, often citing concerns that the race for profit is outrunning the commitment to caution.

So…

When one of the brightest minds in AI safety decides that writing poetry is a more effective way to help humanity than building guardrails for Claude… we should probably pay attention….. right???

Or has old mate just taken his fat pay packet and decided he can’t be arsed with the office BS anymore?

Either way – respect for following your human calling.

I have certainly had many moments in business where i’ve wanted to leave it all to create art too.

What are your thoughts on this?

#AI #ArtificialIntelligence #Anthropic #TechNews #EthicsInTech #ClaudeAI