AI coding agents dramatically accelerate software development, with some teams reporting 10–100x throughput gains.

Stripe tracks the number and percentage of pull requests created by its internal AI agents ('Stripe Minions') as one of its most important corporate metrics.

“How many PRs and what percentage of our PRs are created by minions.”

Coding will be a completely solved problem by the end of 2026.

“We're 6 monthsish or towards the end of the year to be completely done with code. Like it's a solved problem and then we'll probably hit some form of, you know, light RSI by end of next year.”

A customer that switched from Sierra to Decagon built 7 new journeys in one month, compared to only 3 journeys built in a year with Sierra.

“Within a basically a month they spun up like seven new journeys on Ducky.”

Poolside's Laguna XS2 was developed from the start of pre-training to launch in only 5 weeks.

“You looked at Laguna XS2 that we launched it was 5 weeks from the beginning of pre-training to launch.”

Jaguar Land Rover went from evaluating 50 designs per day to 1,500 designs per day using Neural Concept AI models.

“They went from 50 designs evaluated per day to 1,500 every single day in production.”

Coding agents have made it exponentially easier to open pull requests, creating a triage and review crisis for open source maintainers.

“It's an arms race where coding agents have just made it exponentially easier for people to open pull requests. And then it's so hard to triage, to review, to understand that it's aligned.”

Within months, dominant mental models around AI will shift away from chatbots and coding agents.

“I imagine in a matter of months, we will look back at today, and we won't even be able to empathize with the mental models that we have right now because they are so over-indexed on chatbots and coding agents.”

Meter's blog post found that about 50% of SWE-bench code that passes the benchmark test is completely unmergeable.

“Meter had this very interesting blog post where they were like about 50% of Sweepbench code that passes the Sweetbench test is completely unmergable.”

An AI-assisted codebase regresses to the quality level of its worst engineer, because that engineer's unaudited patterns get cemented into the code and are then amplified by the AI referencing those patterns.

“The meme that I have is that your codebase regresses to your worst engineer because that engineer who is, you know, very gung-ho about AI and is not auditing their code, their pattern starts cementing into the code and now the AI is referencing their patterns.”

Box does not claim 10x productivity gains from AI for its engineering team.

“We don't claim that it's a 10x productivity game to our engineering team.”

Coding models are approaching a performance plateau, making fine-tuning a viable strategy for use-case-specific optimization.

“We're approaching a certain plateau in how good coding your data to fine-tune a model specifically for your use case.”

In an RCT by Meter, software developers allowed to use AI were actually slower at completing tasks than those not allowed to use AI.

“Meter came out with an uplift RCT, which I think was the first of its kind, or at least the largest and highest quality, where they had software developers split into two groups. One group was allowed to use AI, the other group was disallowed from using AI. And they studied, you know, how quickly those developers solved issues, like tasks on their to-do list. And it actually turned out that in this case, AI slowed down their performance.”