This person is so dangerous that if I offer them to stay in my shaded yard in the middle of the excruciating sun, they will demand that I let them take my house as well.
I don't think the prompts that the author has proposed will actually work. Including final scope and non-scope is good but it's more of a reaction of what the AI already did. These prompts are suitable for a rewrite, basically, since it's unlikely anyone would have had these ready when they start out.
I have found small iterations to have the best results. I'm not giving AI any chance to one shot it. For example, I won't tell it to "create a fleet view" but something more like "extract key binding to a service" so that I can reuse it in another view before adding another view. Basically, talk to the AI as an engineer talking to another engineer at the nitty gritty level that we need to deal with everyday, not a product person wishing for a business selling point to magically happen.
On the positive side, LLMs are trained based on real data so the default is for it to tell you what data showed. Companies will certainly enforce their influence but it's extra effort against the enormous amount of data, just like with trying to censor sensitive topics. Any context used for ads means less context for the user to use which in turn negatively affects their usefulness.
We have seen this before. Companies using VC money to take over the market and then increase prices. In the end, we're worse off without these scumbags but some will still sing that we got free service do it's bot enshitification.
I understand that most people want to move to other more modern tools, it's up to you. However, what baffled me is why the author's choice not to move is a problem? Did we pay them to move and they did not move as promised? Was there some crowd funding to move that was not fulfilled?
I use DuckDuckGo Email and it generates unique addresses that I can both receive emails (obviously) and reply to from that email. There's also an option to shutdown that address and never receive spam again.
LLMs have randomness baked into every single token it generates. You can try running LLMs locally and set the temperature to low and it immediately feels boring to always have the same reply every time. It's the randomness that makes them feel "smart". Put it another way, randomness is required for the illusion of intelligence.
If you already have Home Assistant running, I think it should be simple. Most of the time you can buy devices with pins already soldered and it's just the matter of connecting them together. AIs are pretty good with ESPHome configs. You can even take a picture so that they can help you identify the correct pins. Some coding may be required for drawing things on the display though.
Totally. Frameworks also make it a lot easier for new team members to contribute. React, for example, makes it a lot easier to hire. Any project with moderate size will require some kind of convention to keep things consistent and choosing a framework makes this easier.
Now look at the cross team collaboration and it gets even harder without frameworks. When every team has their own conventions, how would they communicate and work together? Imagine a website with React, Vue, Angular all over the place, all fighting for the same DOM.
Although this is quite a dark time, I hope ICE may finally (accidentally) do some good by making it super obvious to people how much online platforms track them.
You're thinking of open source projects that only need a few hours of work per week. Anything more than a few hours a week either requires someone to give up their full time job to work on it (switch to part time jobs/consultation is an option), or having multiple contributors which still require significant effort to coordinate at the end of the day.
Let's say Tailwind CSS gives up and stops the project, do you think there'll be someone else picking it up, knowing how Tailwind failed in the first place? LLMs don't create new things, they remix what are already available. It's delusional to think that you should use LLMs to create a whole UI library just for your application and spend enormous effort not only maintaining it but also train new team members to use it down the road.
Open source is charity, it's unreasonable, even entitled, to demand someone work on it full time without pay.
Same here. I developed desktop applications for 15+ years and was really frustrated with Microsoft's direction for the UI around the time of Windows Phone. While Windows Forms may not be the best, it worked for decades until then. Now even if someone wants to build a desktop application using native UI, it's next to no resource at all because it's all about cross platform nowadays.