muzani

There's a lot of AI training jobs that fit this perfectly now. Right up to the dissatisfaction - you really do feel like a cog in the system. It pays well, but it reminds me of those stories about Amazon workers who are afraid of taking pee breaks.

They'll promise that the love you and you'll do well, and suddenly you're flagged for low performance because someone used AI to check your work and nobody checked whether the work-checking AI is valid.

Let's say your skills are not up to par. It might be the truth. What next?

There's cognitive dissonance going on here - you are not worthy but you need the money so there's no choice but to be worthy. Your brain has both of these and it's freezing up. That's the part you need to fix.

I might not be worthy either. And possibly not even due to any kind of rot, but because all these brilliant people are competing for the same jobs and willing to do it for cheap.

Personally, I took up entrepreneurship because it's a lot easier to get funds these days and it's safer than all these layoff cycles. What's the point of a full time job if job security is a thing of the past?

I haven't seen it happen. They're the bottleneck now, and for some people, removing the bottleneck may mean removing the process. But there's probably a good reason why it's the bottleneck.

Some people think that as long as it passes all tests, that's all jolly - but it still needs a code review to check whether all the tests are still relevant and whether everything that needs to be tested is tested.

Even if AI gets the PRD 100% right, the PRD may be wrong. Getting the communications right is the essence of software engineering in the end.

GPT and Claude were able to do different takes in the past, but lately they've been horribly mediocre.

There's something called the oatmeal problem in the game industry: If you created enough procedurally generated content, it would become "oatmeal". Incredible variety individually, but bland when mixed together.

I'd say as data increases, it will default to being really bland.

Grok was never the edgy one. It has no competitive advantage so Musk tried to market it that way.

Both Claude and ChatGPT were able to go far more edgy. Claude was a popular smut generator until that time when the US gov locked down Fable, and subsequently it was lobotomized, and ChatGPT was edgy until someone committed suicide on its advice.

They're valued at a trillion dollars faster than any other company in history. I think selling tokens as a strategy works.

They do have plenty of AI products too - RAG, embeddings, voice, companionship, image generation, video. They're not the best in most of these fields but it works well enough.

ChatGPT is also a very powerful platform that they haven't been able to fully leverage.

I love the curiosity despite the possible legal consequences. It's in the spirit of Hacker News.

The flappy bird of our generation.

If it feels vibe coded, chances are that they haven't run it through sufficient QA or a mindful design approach yet.

Or they simply haven't developed the taste to understand why it feels off, and likely aren't domain experts either.

The cynicism here surprises me. Many managers are open to feedback.

But if they're not new to management, chances are that someone has given them the feedback.

And if they're still doing it after a while, that's a sign that they aren't open to it.

I bought a whole bunch of dogecoins in the past, back when it was actually a memecoin. Meme tech communities were awesome. I got tired of crypto at one point and turned most of it to cash, and the rest to DOGE.

Then it actually went up and the whole memecoin space was taken over by grifters.

There's a delay to these things; the value of these communities went up because of the community and it collapsed because the community was taken over as well.

From the outside, looking at all these graphs, it's hard to link them.

Bored Ape was about some celebrities just having fun with technology. It was funny and the music videos were fun.

Survivorship bias.

I was talking to someone about game development this morning. The scheduling to make a simple commercial indie game five years ago was roughly 20 hr market research, 10 hr gameplay prototyping, 10 hr art prototype, 20 hr vertical slice, 80 content & level development, 60 hr on polish. Plus some more on marketing.

Now today with AI, what has changed?

Those who follow a process may instead be spending 5 hrs on market research, 3 on prototyping, 30 on levels, and 100 hrs on polish.

These games are a little better and less painful to make. But the output is roughly the same as before.

Those who don't follow a process will spend 15 hours prototyping, 10 on polish, then find that nobody buys it and blames distribution.

But back in the past, these same people would have never reached a release point. The output of slop has multiplied enormously.

Just being able to give to someone. Children and pets are the best example for this. No matter how badly someone distrusts humans, they can at least rely on pets to give to.

A tribe is a group of people who act as a single unit instead of multiple individuals. A marriage is the same with two people.

Love is the unconditional, unequivocal commitment to the welfare of another. This makes it very susceptible to abuse - you can simply use a loved one for endless money, sex, or labor for example.

Much of society has formed to protect individuals against the abuse of love, often to the point where we think love is evil. Restrictions on marriage, people are decoupled from their jobs, welfare is villainised. Unbounded love leads to tyrants and child abuse.

But love is still necessary for society to thrive. Love is by nature unbounded. And one of the hardest optimization problems of being human is deciding who to love.

Yes, ironically most of it for homework on the days the school is physically closed. Sometimes I need to sign or stamp a document and it's easier to print and scan. We use the scanner more.

Having one where you fill in the ink rather than buying a whole RGB cartridge has been a great decision. I used to replace a whole cartridge up just for reds.