2 ms·

I really don't get the bearish on AI folks. One company has a tool that can replace what we professionally do. This app would have taken person months to buil

I really don't get the bearish on AI folks. One company has a tool that can replace what we professionally do. This app would have taken person months to build, yet here it is, almost as an afterthought. The next ten years are going to be wild. Hope everyone has a plan for leveling up and riding this tsunami. Even if LLMs presently can't do intricate things, that doesn't matter. I know in my gut that my generous total comp job isn't going to be available in the future in its current form. In aggregate, the work we do isn't going to be as hard or as in demand as LLMs devour more of the market share and growth of the human economic field and attention economy. The LLMs are likely going to increase demand for software, but in doing so they'll eat more of the market share. ZIRP, Section 174, industry-wide layoffs, offshoring, and now this. Plan accordingly. I'm sure there are gradients with tremendous success to be had, but for many there will be a decrease in fitness.

hooverd 1y ago
It's that extra 20%. Everybody right now can do the hardware store and rewire their house for cheaper than an electrician, yet they'll still hire one to do it. That or maybe everyone will just except things being slightly more convenient but worse.
- echelon 1y ago
  > That or maybe everyone will just except things being slightly more convenient but worse. This is how it will play out. You'll see it in the cost centers. You'll see it in the "I have an idea for an app" pitches. Business people will make these decisions, and consumers will too. { High convenience, High quality, Low cost } - pick two. Software engineering won't evaporate, but there's something new competing with it.
  - jaynetics 1y ago
    Yeah. I think the analogy with electrical wiring is quite illuminating. Few software errors have a similar risk profile as faulty wiring. Maybe if a calendar or messenger or such is buggy, that's enough of a nuisance to switch, but I doubt a lot of software is this critical. People already put up with plenty of bugs e.g. in games or social networks. Even with paid software, I'm sometimes surprised how much patience customers have developed when it comes to glaring deficits.
    - hooverd 1y ago
      I'd be pissed if my washer or dryer or furnace had the same defect rate as any of the software I used. Software "engineering" has an amazing tolerance for oopsies.

hooverd 1y ago

It's that extra 20%. Everybody right now can do the hardware store and rewire their house for cheaper than an electrician, yet they'll still hire one to do it. That or maybe everyone will just except things being slightly more convenient but worse.

echelon 1y ago
> That or maybe everyone will just except things being slightly more convenient but worse. This is how it will play out. You'll see it in the cost centers. You'll see it in the "I have an idea for an app" pitches. Business people will make these decisions, and consumers will too. { High convenience, High quality, Low cost } - pick two. Software engineering won't evaporate, but there's something new competing with it.
- jaynetics 1y ago
  Yeah. I think the analogy with electrical wiring is quite illuminating. Few software errors have a similar risk profile as faulty wiring. Maybe if a calendar or messenger or such is buggy, that's enough of a nuisance to switch, but I doubt a lot of software is this critical. People already put up with plenty of bugs e.g. in games or social networks. Even with paid software, I'm sometimes surprised how much patience customers have developed when it comes to glaring deficits.
  - hooverd 1y ago
    I'd be pissed if my washer or dryer or furnace had the same defect rate as any of the software I used. Software "engineering" has an amazing tolerance for oopsies.