[bradleyjg]

> That's a _huge_ shift. Most people I know cite +20%-40% velocity with these tools, against the actual work their company cares about doing.

We all have our own observations and mine don’t significantly diverge. But that’s bottom up. At this point shouldn’t we be seeing it top down?

If we are beyond potential and into significant productivity gains, why isn’t that showing up for the customers?

Why didn’t delta airlines get significantly more operationally efficient in the last 3 months due to the introduction of better software?

This is a genuine question, I am seeing a disconnect.

[narnarpapadaddy]

Anecdotally, my take on this is that biggest value lever is strategy and alignment, not implementation. The typical company is dozens of little vectors pointed in different directions, and they cancel each other out. Scaling up the magnitude of each is still net zero.

I was recently consulting at org where two separate engineering teams were all in on two different, incompatible deployment platforms and using AI to accelerate adoption of each.

Management was mystified why their engineering leads kept telling them they couldn’t deploy a complete implementation of their solution.

[simonw]

> Why didn’t delta airlines get significantly more operationally efficient in the last 3 months due to the introduction of better software?

The coding agents got good in November. Most individual engineers didn't fully clock this until January/February. This means that companies didn't really figure it out until March/April.

Assuming companies like Delta have adopted coding agents (which would be pretty fast) it still takes months from adopting a new tool to the code results of that tool rolling out to production.

I expect (and would hope) Delta's software development culture is very conservative. Since nobody can confidently tell Delta "here are proven practices for using this tech to produce high quality, more secure code" yet it would be surprising if they were blasting full-steam ahead.

I expect that even companies that got on board with coding agents in January will only just be starting to ship user-facing features that benefited from those new tools. Shipping software takes a long time, no matter how much faster the "typing the code in" bit gets!

[nchie]

>The coding agents got good in November.

Maybe irrelevant to your point, but I'd argue they were really good already in May if one used the right workflow (planning etc.). They've become better, but they're not saving me significantly more time now than they did 12 months ago.