Abhishek Shankar's Blog

Three startups building AI mathematicians raised over $580 million in the last twelve months.

Here is a claim that does not ask you to trust the mathematician who posted it, the AI lab it credits, or a word of the coverage around it.

The most cited number in enterprise AI is that 95 percent of corporate generative-AI pilots deliver no measurable return.

Anthropic shipped Claude Sonnet 5 today.

I argued on this site that the harness is the product, that the frontier model is a wasting commodity and the thin layer of scaffolding wrapped around it is the asset you actually own.

Google Cloud shipped the Open Knowledge Format in June, and the reaction told you everything about what the agent-tooling world believes its real problem is.

The most-installed agentic coding tool I have seen this year is not a model, a framework, or a runtime.

For two years the reflex when an agent failed was to reach for a bigger model.

When a gatekept industry looks vulnerable, the instinct is to declare it doomed.

A six-word sentence put the AI-coding timeline in a chokehold this spring.

Every benchmark you have ever seen measures correctness.

The story everyone tells about the AI race is a story about intelligence.

The math breaks because the headcount never moves The cleanest articulation of the problem came not from Uber but from a chip company, which makes it hard to dismiss as the complaint of a firm that planned poorly.

I have spent the better part of a year building agents that are supposed to get better on their own.

The framing most teams reach for when they start building agentic workflows on top of a content management system is architectural: the CMS is the database, the agent is the application layer, and the integration is a matter of wiring the right API calls.

On May 20, 2026, OpenAI published an eighteen-page PDF containing a proof that disproves a conjecture Paul Erdős posed in 1946.

The autonomous-research-agent genre stopped being a curiosity about eighteen months ago.

The debate over how to make agents persist has hardened into two camps.

There is a way to read the literature on class-conditional flow matching that makes it sound like a minor convenience.

On May 14, 2026, an APK teardown of Google app beta v17.23 surfaced a feature with a name nobody had heard before: Gemini Spark.

The number people are reacting to is $1.5 billion.

The dominant story about subquadratic architectures for the last five years has been that they were coming for the transformer.

Services firms do not sell skill.

What's missing from today's alphaXiv trending feed is more telling than what's on it.

The $5,000 AI mini PC market has a quiet structural problem.

Every agentic AI company talks about reasoning.

A paper called Weak-Driven Learning hit 245 upvotes on the Hugging Face trending board today, which makes it the most-upvoted paper on there right now.

Every authentication system in production right now assumes the same thing: the entity holding the credential is the entity the credential was issued to.