I agree that imperial units are silly and we should probably run on metric for lots of practical reason. But you gotta admit that they're a fun discussion prompt for trivia; everybody thinks there's an internal logic but it's just all weird history! For example nobody's gonna believe that it's just a weird coincidence that a Zacharia is slightly *less* than half of a Scaramucci
trying to think about the defining characteristic of the entire Zork™ franchise and it's such an interesting microcosm of the world of IP. When you try to perceive a franchise as an artistic expression rather than a financial relationship. What makes a Zork™ a Zork™? Well, it's a game you can play on a computer, that is either silly or serious, where you are or are not underground for all or some part of the narrative
Apologies; I boosted an article earlier from morewrong dot lol, which appeared to be a savage parody of lesswrong. Which, I guess, it is. The header image looked a *little* slop-y but the writing seemed like a dead-on parodic voice.
I guess this is just because it was slop and the lesswrongers have trained themselves to write like slop? In isolation I don't know that I could have clocked it; having perused the site a bit more I see that it's obviously a bot repeating the same joke over and over
this does seem to encode a lot of … I would say "falsified" waterfall-era ideological assumptions about how software development works, like you can measure "delivering features" separately from "bug fixing". and yet it routinely relegates humans to "bug fixing"
some salient takeaways before a deep dive, though:
- it's a … monte carlo simulation? where they're comparing against a "human baseline"? how does *that* work
- on average, 61% of the agent budget is for bug fixing
- "at realistic velocities, most money is spent on developers"
- "Monitor technical debt, since agents tend to increase it"
- "When agents are capable, developers spend 88–90% of their coding time on bug fixing, while few (if any) agents are used for bug fixing" wow, *miserable*
this seems to be an interesting result although the math is far beyond me. I'm definitely curious to hear if anyone looks at the methodology for this. it does seem to indicate that you can deliver more features at lower cost with agents, and I don't want to reject that out of hand
Also this is a big argument in the alternative. For technical reasons related to context limits and training data, none of this shit actually works well enough to replace software at any meaningful scale. To paraphrase @pluralistic by way of Jensen Huang: AI will not take your job. A boss using AI to fake their quarterly KPIs will take your job.
This is an attempt to address the growing number of "oh, so you don't like automation when it comes for *you*? checkmate, programmers" articles out there, but there's so much more to say. Overall I think this idea that computer programmers just destroyed jobs is largely false. One of the weird corollaries of the Productivity Paradox is that we mostly *did* expand possibilities and make things better and more fun.
(Until all the Gig Economy stuff. But that's still a minority of tech history.)
AI does not disrupt. AI entrenches. AI turns money into more money; AI turns investment into dependency. We are living in a funhouse mirror of that right now where the combination of overly-plausible demos and first-one's-free incentives are swamping the actual balance-sheet math. But the economics require that this stabilize, that the capital flows reverse direction. The wave has come in and we are just starting to feel the undertow.
Right now it looks like AI can do "disruption" because you can apparently do things that require a team, as an individual, by taking advantage of highly subsidized token rates, or (at best) taking advantage of highly subsidized training regimes and stolen training data. But those rates will go up and while you'll still be able to nominally command those resources as an individual, doing so will still require the capital flow required to hire a team.
Linus is saying the quiet part out loud: AI-native projects are environments where the key ingredient is _capital_. Abstraction can be performed by a quiet appreciation. True automation can be disruptive.
AI-based architectures (again, assuming that the AI works reliably, which is a highly contingent proposition) translate all that stuff into capital flows. A "machine speed" project requires _money_ to participate in, regardless of their complexity.
Abstraction is leverage. Automation is the amplification of power. Abstraction is democratization. Automation repeats a duplicative process so you can stop thinking about it and focus on something higher level. Abstraction elides so architecture can expand.
AI is anti-leverage. Where an abstraction will elide duplication with semantic compression, AI steamrolls the duplication by copying and pasting faster than a human can.
It's also very frustrating to constantly have my opposition to AI be taken as an opposition to abstraction, or to automation, and a desire to do every little thing by hand. But it's the opposite. For a whole bunch of reasons, AI anti-abstracts. It adversarially creates oceans of duplication to process. It creates a situation where the AI user gets to set the terms of some ecosystem with synthetic productivity but then everyone else needs an AI to figure out what's going on in that new ecosystem.
Read the _entire_ thread, please: https://cosocial.ca/@mhoye/117439531864946276
Linux is lost. In the meanwhile, we'll be stuck with it for a decade at least, but it's gone. I have no hope that anyone involved is capable of learning or course-correcting any more, and it was badly off-course before the slopocalypse. This is just the final nail in its coffin.
Which means that there's a good chance somebody reading this is going to start the project that becomes its successor. When you do, *learn these lessons*.
(It probably helped my resolve that I saw a very young friend develop a very long-lasting disability from typing-related RSI quite early on, which impressed on me that this could spiral out of control extremely quickly.)
Now, I am accommodating to a fault. I have always taken my professional obligations _very_ seriously. I do not like to push back on management especially if I understand their reasons, and "we have a deadline beyond our control" is absolutely a reason I understand.
But "this is hurting me, and if I keep going it will hurt me for MUCH longer" has always been a real bright red line for me where I was willing to quit on the spot over it.
You'd think that this would be a simple proposition. My job is injuring me in an observable way, I don't want that injury to persist. Time to stop; there will be more work tomorrow. This is the kind of thing that union rules or OSHA regulations enforce in a lot of professions.
Not programmers! In more than one case I have had to tell a manager "my wrists hurt so I'm out for the day, see you tomorrow", and they'd tell me that I couldn't stop working.
I _have_ compensated for this bias in a few key ways. One of the issues on that list I have *not* experienced is RSI. A big reason for that is just luck. But another reason is that I'm sensitive to wrist pain, and when I notice wrist pain from typing, I stop typing. Even if I have a deadline, regardless of what else is going on.
I probably need to be clearer here because I am not exhausted because this is a stupid bias to have. I have this bias. I have this bias EXTREMELY strongly. It is very easy to acquire and extremely hard to compensate for. I have experienced a significant share of the health issues from that laundry list.
But that's what makes it exhausting. My brain is already tilting the floor super hard in this direction. That's what makes it extra frustrating when someone else tilts the floor even harder.
I'm talking about AI of course but this is barely even a problem that AI has made worse. "Sure, I'll think about the problems, but it's just an app, what harm could it do" says a programmer under 50 with sciatica, myopia, diplopia, hypertension, repetitive stress injuries in their neck and wrists, heart disease, chronic anxiety, chronic headaches, burnout, depression, and cPTSD