wasabinator

Props to the author, this is very well written. Makes a lot of sense.

I have seen first hand that AI is turning senior devs into junior devs. The easy option of just vibing things out is just too seductive for most, and I see now most PR comments "from" reviewers are just copy pasta (or posted directly by) their agents. This is not a good way to train your brain, and it feels like there is less and less desire to seek correctness instead of just passing code through more rounds of AI review, which by virtue of how LLMs work never gives you the same result any pass.

People used to know in very clear detail what they just worked on, what key decisions were made, and why they were made. The memory would hold for at least 3-6 months and longer with a quick refresh. Now people hardly understand in detail what they pushed out just yesterday.

Think of how many times you have been asked "oh yeah what did we do about x in the past before?". Now that question will be greeted with blank stares. "I'll ask Claude".

> People are saying the pace will settle down once LLMs find and fix all the bugs

Yeah I don't buy it either. Look at github right now, it is a total slopfest with people vibing projects they don't know the first thing about. Plenty of LLM use will be delivering code full of new exploits. It will be interesting to see the frequency of these exploits being found correlated to LLM friendly/unfriendly contribution policies once the dust has settled to tell for certain though.

Yet more language bloat

Scrum has always been a waste of time attempting to solve problems that it created in the first place. All the planning burden of Scrum invariably goes out the window the moment you go to build something and realise every assumption was incorrect.

Then there is task sizing "oh it's not a time estimate" they say, then they schedule n points per two weeks, with the work sized in "not a time estimate" points now literally converted into a time estimate on the team's behalf. In addition, it then gets phrased as a "commitment" by the team of what will be delivered.

Kanban is really the only honest form of agile, with no wasted upfront planning commitment. Take top card, move it to completion, repeat. Prioritisation moves the card order and it doesn't have to be the devs job, more a product person can decide that.

I would say it could be less lying and more the various "authors" having not even the slightest idea what their creations actually do.

2050 the future of Linux in the pocket!

AI;DR

I saw the board cost, but how about the total cost in tokens to design it? Can empathise with the good and meh feeling you got from it. For me I try to stick with manual struggles with new concepts which grows my brain and allows me learn something new. But its impressive what LLMs can do.

It's less hating coworkers and more enjoying the perfect logic that comes from interacting with a cpu versus the fuzzy nature of human moods etc. It's likely why a lot of people (myself included) got into coding.

The current token pricing will not continue forever (capitalism 101), it's a bait and switch. Let's see how cheap this all is in a year's time.

When they use ill gotten data for training their models the world's smallest violin plays when they complain about distillation attacks.

Yes but the point of the article was simply that the tech was 10 years ahead of it's time, but it provably wasn't.

I'm an Amiga diehard from the start but the industry caught up in a lot less than 10 years. Also, the amazing X68000 was a thing in Japan as of 1987.

Do we still fall for the China = must be spying, US = wouldn't spy unless it's disclosed?

It's fair to say whatever software you run from any country of origin is hoovering up as much data about you that it can get away with.

If you haven't compiled open source, audited code and are running on that, all bets are already off.