How much value is AI creating? | Hacker News Reader

How much value is AI creating?

ft.com

The article is referencing this paper, btw: https://www.nber.org/papers/w35275.

As I commented on the Ladybird thread (https://news.ycombinator.com/item?id=48410252), this is an empirical result that sort of tracks the intuition behind "sending a PR that compiles is no longer a sufficient stake to show real commitment."

I think this article echos a lot of what people have been saying here on HN, that there is a massive funnel of productivity.

The article says that whilst the number of edits to individual files has increased by 300% actual releases have only increased by 30%.

I even find this 30% seams higher than what I am seeing.

I love the quote by the Japanese engineer Ohno that a broken machine or system makes a lot of noise. I think it sums up the current situation precisely.

And archived https://archive.is/5nHAP

A really interesting point in the article that even though the number of IOS apps Has increased by 80% the number of reviews has actually decreased.

The number Apps with significant usage has gone down slightly.

Is this suggesting less critical thinking?

I think those are unrelated dynamics. AI isn't causing people to use fewer apps. (Well, it might be, in a different way: you don't need the Yelp app if you can just ask ChatGPT where to get lunch.)

It's rather the case that AI makes it easier to build and publish an app, but not substantially easier to make a popular app that people will actually use.

Yes, I agree. There are two theories here Quantity has gone up but quality has gone down, or people need less apps as the LLM answers their needs.

Trying to answer this is important to try to answer the future. What startup to build

It's not surprising that any new transformational technology (from railways to electricity to computing to internet and beyond) did not immediately create a productivity boom.

What IS surprising (or rather, infuriating) is the insistence (from various stakeholders, including, of course, the purveyors of said technologies, and their multiple investors) that productivity show up IMMEDIATELY, or at least in a relatively short time frame.

THAT is what this article addresses, quoting the NBER paper https://www.nber.org/papers/w35275 that shows empirically the AI-driven productivity (or its lack) in SW development. It shows that while edits and PRs increased 300% (in the study), the actual SW releases only increased 30%, and the subsequent usage of those AI-developed apps did not increase meaningfully at all. Hence the hand-wringing over whether AI-driven value creation exists or not.

As with any tech deployment, productivity boom really relies on *transformations* of orgs, org structures and business processes, and not merely by sprinkling the magic-sauce-du-jour. This is a well-known and well researched insight, and follows from various examples (e.g. introduction of computers in '70s and '80s for accounting, finance; industrial robotics; electrification; mass communication; etc)

Why would AI-driven productivity be any different? Why would anybody expect IMMEDIATE payoffs? Real payoffs would take years, if not decades, to show meaningfully. [This is what infuriates me, tbh]

Until then, orgs will do what they have always done. Reduce headcount, bank the short-term gains, and claim "See? We are AI-native! Gimme more valuation!"

Reply on news.ycombinator.com