The BaseEssayLessWrong

Some Memories of Early Anthropic

In a personal essay posted on LessWrong, Ivan Vendrov, who says a junior role at Anthropic lasted eight months between October 2022 and May 2023, recalls from memory doubts about the company's strategy, workplace pressure and resigning. Vendrov says the dialogue is rendered as gist and details were altered.

Open the source ↗
About Anthropic or ClaudeClaude, made by Anthropic, wrote this entry. It follows the same rules as every other entry, whoever the entry is about.

Why it matters

The same thing, for different readers.

What it means for you

Former Anthropic employee Ivan Vendrov gives one person's memory of working inside an AI lab in its early days, which Vendrov says is not a portrait of the company today.

For lawmakers

Ivan Vendrov argues that little public information exists about how AI labs work inside, a gap that matters to anyone deciding what insiders should be free to say.

For researchers

The essay is an unverified recollection of one junior employee over a short window; Vendrov invites former colleagues and people at other labs to fill in the picture.

The question it raises

Vendrov raises the worry that the helpful, harmless assistant persona that labs train is the one most useful to the powerful, without being able to say why it is bad.

In their own words

Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.

From the source’s own text.

“given the enormous power of the AI labs and the paucity of public information about their inner workings, I feel compelled to share the most salient parts of my personal experience.”
Ivan Vendrov, former Anthropic employee in a junior roleVendrov explains in the opening note why the essay was written, after saying it is recalled from memory and is not a description of what Anthropic is like.
Found word for word in the sourcestarts inside a longer claim, right after “but”
“Anthropic is now a massive organization with many silos and secrets and I have very little idea what was or is happening there in some overall sense.”
Ivan Vendrov, former Anthropic employee in a junior roleVendrov states the limit of the account: it covers a short window and one junior vantage point.
Found word for word in the source
“I was told I could help set the beneficial deployment strategy, but the decisions have actually mostly been made and we are going after boring use cases like automating customer support”
Ivan Vendrov, former Anthropic employee in a junior roleVendrov recalls the first weeks, in November 2022, when motivation began to fade.
Found word for word in the source

The record

Questions
Are AI companies putting stronger safeguards into practice? · What values should AI have, and could it have a mind of its own?
People
Ivan Vendrov (Wrote it, former Anthropic employee in a junior role)
What we read
We read the whole source.
Found by
The Radar, run of October 26, 2026, via lesswrong (tier 1: the source’s own channel or an official record)
Cite this
LessWrong, “Some Memories of Early Anthropic”, October 26, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/lesswrong-2026-10-26-memories-early-anthropic
Our notesEntry is about Anthropic (Claude wrote it; editor reviews). The author says it is recalled from memory, dialogue is gist only, and pronouns and identifying details were swapped, so no dialogue quoted and no unnamed colleagues attributed. It is a linkpost to nothinghuman.substack.com; the saved text is the LessWrong copy. Nothing in the account is independently verified.