In my experience, LLMs are great at reviewing changelogs for potential gaps from a user POV (and even creating draft changelogs wholesale, if you're backfilling) based on git history.
> So doubling isn't because it is doing above everyone's expectations, it is because Apple underestimated the demand.
Clearly it's doing above their expectations, and they had precise data in the form of their test selling the M1 Macbook Air at $599 (occasionally $499) since 2024. It's too bad you weren't at Apple so they could've avoided this mistake!
> …"progressive web apps", originally coined by Google.
To make it a bit less faceless while humans still matter, it was coined by Alex Russell and Frances Berriman in 2015. Russell worked at Google at the time, Berriman at Code for America.
My perspective: I've been supporting and working with Mac and iOS developers since the last century, when Apple moved me from Chicago to be an evangelist at Apple HQ in Cupertino. I know as much as anyone about AI-assisted app development, as the creator/maintainer of the free/open source Axiom (https://charleswiltgen.github.io/Axiom/) for iOS/macOS devs.
It's not as dire as you might think. To software developers, the "AI revolution" is largely what the "desktop publishing revolution" was to designers. Yes, it meant the "riff raff" could theoretically play with the pros. Some percentage of the riff raff became pros. Most of the pros eventually adopted the tools and techniques used by the riff raff. Some of the pros didn't survive the transition and happily retired, taking their rubylith, Letraset type, and rubber cement with them.
The silver lining is that most of a software engineer's job isn't coding, it's thinking. LLMs can't do that, and we're not getting to AGI with current AI architectures. LLMs can amplify thinking, and an LLM in the hands of a software engineer or architect is at least two orders of magnitude more effective than it can be in the hands of a vibe coder. As LLMs get better for vibe coders, they also get better for pros.
One can argue that, by the end of the decade, hand-coded [your language of choice] may be considered as unnecessary as hand-coded assembly has been for decades. But coding in modern languages is already 7-8 levels of abstraction above the metal. One more level of abstraction is not the death of software engineering, IMO.
I don't know? It's an agent, an automated contractor, a black box that produces work. When you pay someone to help write your app, you don't call that an "abstraction."
I think the reasonable way to look at LLMs is similar to how you'd work with junior developers. They can churn out code, but do need constant guidance. Of course, (some of the) real juniors will eventually become seniors. It remains to be seen if LLMs will.
I don't think i fear so much for competing on quality vs a vibe coder, I think two orders of magnitude is underselling the potential snowball effect of a vibe coder attempting to craft up a semi complex project. The sooner they start to pile on complexity, the quicker it is that the 2x gap widens.
The fear for me comes with that initial creation, which is a fear i have across a large amount of things for AI "disruption". Just floods of bad products at rates that are insane, same type of problem that exists on github of just garbage prs, and same thing for any reddit/twitter/even hn type comment thread.
We've banned this account. Posting repetitive, snarky, low-substance comments like this is clearly against the guidelines. If you don't want to be banned, you can email us (hn@ycombinator.com) and commit to observing the guidelines in future. https://news.ycombinator.com/newsguidelines.html
Admittedly, I’ve gone through waves of feeling scared for the future of software engineering. I manage many people so I care about them and their futures. But the more I use AI in writing software, not just vibe coding, the more I align with this perspective. And the way I see it, when AGI does come it’s going to affect a whole lot more than the software engineering job market. That’s global societal-level impact type stuff. We’d have to reckon with if work (meaning a 9 to 5) is truly a requirement to thrive in life. Spoiler, it isn’t.
> It would be extremely cool to be able to write one or two lines of prompt in my harness, and have a light model iterate with me a few times writing/proposing requirements, guidelines and explanations, refining the prompt until it's ready to be sent to the actual LLM.
It is cool and (IMO) necessary, and most AI-using coders I know do this using skill suites like Superpowers (see: /superpowers:brainstorming). https://github.com/obra/superpowers
> The trick is to have a fresh instance doing the reviewing, not the one that did the work.
In my experience that's not neccessary (some people even claim that you must use models from different vendors), and it's expensive since a fresh instance needs to rebuild all the context that's needed in order to properly and thoroughly review. LLMs have no problem throwing "them 5 minutes ago" under the bus when asked to review something "skeptically" and "with fresh eyes".
Doing it in the same session does save a ton of tokens but I find it's too biased towards its own implementation even if you tell it to use "fresh eyes" or to "act like a code reviewer in a bad mood." Including those strings in your prompt does show some improvement but not nearly as much as making it think from first principles in a fresh instance.
It's a bummer that Wi-Fi Alliance completely fumbled basics like standards for naming (Miracast has 30+ names) and UX consistency. It's not much of a mystery why Apple is throwing its weight behind Matter Casting and even Google Cast.
In my experience, LLMs are great at reviewing changelogs for potential gaps from a user POV (and even creating draft changelogs wholesale, if you're backfilling) based on git history.
reply