Philip Crawford 🌐 🏗️ 🚲
banner
philip.health
Philip Crawford 🌐 🏗️ 🚲
@philip.health
🐈 🐕 cat and dog fosterer
🚞 train traveler
🌐 neoliberal
🏗 yimby
🚴‍♀️ bicycle commuter
🥏 Ultimate Frisbee GGM

Friends and co-workers in 🇺🇦 so I follow that invasion closely.

✅ www.manifest.ly
Yeah man, you better short MSFT and GOOG asap! 😂

You are a lot dummer than you think.

en.wikipedia.org/wiki/Dunning...
November 28, 2025 at 10:42 PM
I'm sorry, but you clearly have no idea how LLMs actually work. If you continue with your anti AI stance, you will become a dinosaur. Good luck with that.

www.anthropic.com/research/tra...
Tracing the thoughts of a large language model
Anthropic's latest interpretability research: a new microscope to understand Claude's internal mechanisms
www.anthropic.com
November 28, 2025 at 10:40 PM
Exactly.
November 28, 2025 at 9:54 PM
That is literally not how they work. They are prediction machines. Why are you so dead set against learning about a new technology?? It's wild to me. Are you older than 50?
November 28, 2025 at 9:53 PM
I agree with that. I'm more interested in how you plan to transition to using AI in your development processes. The models and the tooling today are **significantly** better than they were a year ago. I expect to say the same thing in 12 months.

So how do non users plan to keep up on the tech?
November 28, 2025 at 9:40 PM
This is objectively not true. LLMs are not copy paste machines. Why do you keep making that claim??
November 28, 2025 at 9:33 PM
True. Increasing returns and path dependency would predict that none of these firms would choose current profitability. It's similar to Amazon choosing to forgo profit for many years.

That doesn't diminish the value we get from using AI.
November 28, 2025 at 9:04 PM
I am curious what your plan is moving forward. Even if the tools don't seem to improve your productivity now, how do you plan to experiment with them moving forward? In my world of SaaS, I would not hire a developer without significant experience using AI. But maybe web apps are different.
November 28, 2025 at 9:02 PM
It is not. JFC
November 28, 2025 at 8:53 PM
Do you plan to experiment more next year using AI for programming? Perhaps they don't benefit you *now*, but they have been steadily improving and it seems it is only a matter of time that every programmer will be using them in an integrated fashion in their daily work.
November 28, 2025 at 7:43 PM
How do you come to that conclusion based on this thread?
November 28, 2025 at 6:55 PM
I don't think you understand how LLMs work if you think they can't invent things. I would recommend experimenting more with the tools. They are not copy/paste machines and they do not need prior examples.
November 28, 2025 at 6:55 PM
If AI is useless for you, that says a lot more about you than it does AI. I'm blown away that any software developer would make a claim that AI is useless for development. The studies I've seen, and the experience of my team, is far from that.
November 28, 2025 at 6:46 PM
We have added a considerable number of features and have not raised our price in several years.
November 28, 2025 at 6:43 PM
That's a great question. We use the OpenAI API a LOT and during the time we've been using it (late 2022) the prices have dropped more than 10x for calls to the models we use. If they increased price 10x tomorrow, we might stop using for some things, but most we'd continue.
November 28, 2025 at 6:22 PM
I'm in the business (@manifest.ly) and can assure you that productivity is definitely going up through the use of AI.
November 28, 2025 at 6:10 PM
Software is clearly being made cheaper by AI and considering how much software is built into the world of services....
November 28, 2025 at 5:52 PM
Amazon is gonna deliver letters for half that price and then lefties will complain about their monopoly and taxing billionaires more.
November 28, 2025 at 4:46 PM