If you need a quick&dirty piece of code, you can generate it by firing some prompts into an LLM. You can get a decent result if you know what you are doing. It may not be production-ready, but often it’s a good starting point. This is NOT vibe coding.

Vibe coding is what people who don’t know what they are doing do when they try the process above, and think it’s production-ready.

Maybe figure out if there’s a less emotional/defensive path for looking at all this

If you start with

I think some people are so hung up on knee-jerk defensiveness of AI that they lose sight of everything but promoting it.

you should expect to be seen as one of those with that irrational hate for a technology.

Back in topic: can this be a bubble like the dot-com? Of course. Is it? Probably not.

95% of failures should be a warning for all those fools who expects something that this technology cannot do. Nothing more than that.

About 90% of tech startups fail. It happens all the time because it’s in the nature of innovation.

Here anything about AI is received negatively so a 5% success rate is the “demonstration” that it’s a bubble. I’m sorry if you hope so, but it’s not. 5% is not far away from the normal failure rate of new companies and here we are talking about early adopters who buy a lottery ticket trying to be the first that makes it work.

Feel free to believe the contrary. I don’t need to convince an anonymous guy on internet.

Like any technology before it, now we are beyond the hype in the area where lots of clueless people expects miracles and complain about the number of Rs in strawberry.

In one year or two it will be a regular tool like any other.

5% to me sounds excellent. Companies fail all the time with well established technologies and nobody should expect great results with something still so new. It’s a bet with high risks and high rewards: most people will simply fail.

Using legit regular banks would be your best bet to avoid scams or being accused of fiscal crimes.

It depends on where the billionaire’s and your bank account are located (some restrictions may apply), but a normal bank transfer works perfectly for any amount. Of course for large amounts it’s better to contact the banks first because it may reject a large deposit if it comes out of the blue.

For donations there will be taxes and some bureaucracy, the anti money laundering will be triggered, but it’s nothing impossible to handle.

Building a filesystem essentially means linking a directory of filenames to physical blocks and handling CRUD operations. It’s not that hard. The hard part comes when you go beyond the basics to build something efficient with useful features. For example, fast access, journaling and fragmentation are all challenging topics. You can try without messing with the kernel by creating an in-memory filesystem (essentially a block of RAM) and playing with the I/O.

I would not hold my breath considering that GH was already supposed to die in 2018 and we are still predicting its imminent end in 2025.

Wrong assumption 👋

Damn, you got me! Now I have to rethink my whole plan to undermine Spotify’s revenues one comment a time!

I never said it wasn’t a successful service. I’m only genuinely surprised that so many people pay for what I personally consider marginal improvements. Also thanks for being the usual internet jerk ready to snap at any comment he doesn’t like: now I have an idea of who may pay for that subscription.

What bubble?

Same with pi-hole.

You guys are paying for Spotify?

I guess I’m the only one who sees this as “here little Donald, now play with this and let me do my job”.

A better resource: https://www.goeuropean.org/

Also useful: https://isitamerican.eu/

Maybe now they have a chance to demonstrate that these filters don’t work.

Plus: shady managing of the network connections that could lead to mass surveillance, poor E2E encryption, and now a partnership with Musk.