First AIE Code Summit speakers:

@poteto, Principal Engineer @ SpaceX
@Steve_Yegge, Engineer, Writer, Advisor
@zachbruggeman, Principal Software Engineer @ Ramp

Nov 10-12, SF. Come hang out with us. Apply at ai.engineer/code

Your coding agent's verbosity is a code review problem.

Ben Holmes at AIE NY on scoring redundant tests and noisy explanations.

Scaling without Slop at AIE CODE.
November 10 to 12, San Francisco.
Apply to attend: ai.engineer/code

Scaling without Slop: Melvin Johnson, Google DeepMind
System One models: Diogo Almeida, TypeSafe AI
The Bun rewrite: Jarred Sumner, Anthropic

AIE CODE. November 10 to 12, San Francisco.
Apply to attend: ai.engineer/code

First time at the Latent Space studio after being three year listener

First time interviewing a true inspirer and podcaster that really explains the AI’s complexity in “human” way

First time we are going to bring @aiDotEngineer to Shanghai with Mengying and her team

Thanks for and can’t wait to see what is coming in Shanghai!

@aiDotEngineer NY videos are like crack for finance/compute folks 🔥

Asking an LLM to write json or a custom DSL is like asking Shakespeare to write a poem in Japanese.

HTML, CSS and JS are their native language. So that's why we chose them for @HyperFrames_

This short was edited in @HyperFrames_ Studio from my @aiDotEngineer talk

join the first AIE in China! 🇨🇳

"If you just said agent extract Excel, extract PDF, you might get good outcomes, but they're not going to be reliable."

@mitch_troy: reliability comes from opinions, and some come from expertise. Read the primary source before answering a tax question.

@aiDotEngineer NYC:

A family turns down an offer for their business. I'd want their agent to remember more than "they don’t want to sell."

Maybe they wanted another year with their team. Maybe caring for a parent meant they couldn’t face another big change. When another offer arrives, those reasons might still apply. Or they might not.

I’m building X1 Wealth, an AI-native family office, and @fyxlong’s "hedge fund as a function" talk at @aiDotEngineer wonderfully articulates how something like this can be approached.

He framed a hedge fund as a function that produces decisions about what to invest in, how much, and when. Then he broke that process into smaller steps, each with its own inputs and outputs, so you can test what actually does the job best: ordinary code, a model, or a person.

For a family office, part of the work is understanding what the family wants its money to make possible. The same sale price could support very different lives. I want an office that understands those choices and helps carry out the work they lead to, with the family’s financial pros involved.

Each decision adds to the history the office can use next time. A better model should be able to pick up that understanding without making the family start over, while leaving room for them to change their minds.

And that history gets personal pretty quickly. Understanding why someone made a decision doesn’t mean everyone helping carry it out needs access to those reasons.

If your agent already knows something about you, you might try asking:

"What are you treating as a lasting preference that might have been a decision I made under particular circumstances? Show me what you’re basing that on, and tell me where you’re guessing."

I’d be interested in whether you recognize yourself in the answer.

Frank’s talk: A Hedge Fund as a Function
ai.engineer/nyc/2026/talks…

We’ve seen people still running Llama 2 in 2026.

Too many teams have no idea what models they’re actually running, what they’re paying for them, or why they chose them.

That’s why our CEO @picocreator took the stage at @aiDotEngineer NYC with a workshop on exploding token bills, and how to get yours back under control.

Full talk below!

youtube.com/watch?v=Ti-BmG…

Only the baker knows what the bakery needs.

@seldo of @arizeai on where the bottleneck moves as code gets cheaper: understanding the business, describing the problem, and defining what a good solution looks like.

“I don’t do bugs.”

We’ll let the code review decide.

@QodoAI at AI Engineer New York.

Photos by Ad Image.

this is what talent density means at @aiDotEngineer - @SydneyScouts runs speakers at AIE!!

this entire conference was awesome. congrats to @swyx and the team for another banger. bringing ai + finance teams together in nyc for what i hope is the first of many... looking forward to aie code summit next month!

Three more AIE Code Summit speakers 👀

@lydiahallie, Member of Technical Staff @ Anthropic
Utkarsh Shrivastava, Engineering Leader, Enterprise AI @ Netflix
@philipkiely, Author of Inference Engineering @ Baseten

Nov 10-12, SF. Apply to join us! ai.engineer/code

My AIE talk is live!

I talk about the benchmark I've been building for 5 months now and released it publicly for the first time there :)

Louis-François Bouchard of Towards AI breaks down what makes AI writing feel like slop: overused words, vague generalities, saying nothing, repetition, hallucination and a voice that sounds like everyone at once. He then shows how his team built a benchmark to measure whether a model can write a video script in his own voice: 10 scripts, 5 drafts per model, 4 LLM judges, nine scoring metrics and 1,000 human A/B checks. He shares what it cost, why judges favor their own model family, and practical tips for writing with AI: very long prompts full of examples, one review pass from a strong model, many cheap drafts merged together, and always a human edit.

Thanks again @aiDotEngineer, @swyx for inviting me and hosting yet another amazing event.

Watch it here:
youtu.be/_dTF3NpPzGo

Here's my take on what a software factory really is

This is my talk from @aiDotEngineer NYC, and I decided to record a version for everyone to watch on X.

tl;dr - software engineering is becoming factory engineering, and there's new work (especially factory benchmarking) that focusing on.

Today is very timely day for this talk from @mkanat at @aiDotEngineer - not SecOps related. But many folks will be wondering which workloads they can pause and stop… and what’s worth keeping around if we can just start fresh on many work streams.

youtu.be/bU5gDVyu_28?is…

Does code quality even matter anymore? My keynote at @aiDotEngineer NYC 2026 just came out on YouTube: youtube.com/watch?v=bU5gDV…

"Nobody wants their agent to move 37 their tax return."

@mitch_troy: review agent work like a PR. Blast radius, and how much you trust the author.

"For agents to diffuse, they need to be simple. For teams to trust them, they need to be reliable."

From @aiDotEngineer NYC.

Laptops open.

Workshop day at AI Engineer New York, October 12.

A few photos of the people on the other side of the podium.

Photos by Ad Image.