apenwarr @apenwarr.ca · 13hI don’t think waterfall is really “slower” in throughput, just iteration frequency. Because you *didn’t* have a working product at each step, there was no way to ship early. But when you reached the bottom of the waterfall 2 years later you often realized you had asked for the wrong thing up top. 130
apenwarr @apenwarr.ca · 19hIf you believe 90% of waterfall projects failed (which is... unfortunately believable), then productivity increased 10x just by iteratively shipping smaller things that had value. 180
apenwarr @apenwarr.ca · 30/09/2026Yeah, that’s basically a chapter of the book. Others include how to find out what it is you want to be true 100
apenwarr @apenwarr.ca · 29/09/2026Well there’s horses. They have been extremely resistant to retraining for office work 100
apenwarr @apenwarr.ca · 29/09/2026You say "if." Is this a pattern you've observed in history, or a hypothetical new thing? 100
apenwarr @apenwarr.ca · 29/09/2026Maybe it's this: in software companies, we always underestimate how hard it is to take a product from "working" to "actually can be sold in high volumes." It's typically years of work! I think AI gets us to "working" really fast. But the rest of the job requires slow real-world feedback loops. 040
apenwarr @apenwarr.ca · 29/09/2026Everything is faster now than it was during the industrial revolution, undoubtedly. But AI will also hit all kinds of speedbumps when it touches the physical world; it already is. Plus modern fixes to problems (reskilling etc) are accelerated as much as modern causes of problems. 240
apenwarr @apenwarr.ca · 29/09/2026Wait, are we talking about employed people spending money on US services, or unemployed people in a country where... where did the local jobs go exactly? 100
apenwarr @apenwarr.ca · 29/09/2026In my mind, AI is a specific technology the way "the concept of factory automation" is a specific technology. it's useless abstractly, but transformative as we apply it to an endless series of specific problems. 150
apenwarr @apenwarr.ca · 29/09/2026That your point about "lower-valued jobs" and "payroll savings" is the opposite of what has been true with every other technological advancement in history 010
apenwarr @apenwarr.ca · 29/09/2026The US is in the middle of losing a trade war of their own making. There has never been a better time to tax US-based goods and services. 100
apenwarr @apenwarr.ca · 29/09/2026If that comes anywhere close to happening, the countries where those people live will figure out how to collect local taxes on the foreign companies. Don't worry. 121
apenwarr @apenwarr.ca · 29/09/2026It really does make a person feel rich to own an actual part of the doomsday machine 110
apenwarr @apenwarr.ca · 29/09/2026Certainly not. That would be unethical. Usenet is for noncommercial use only 0371
Reposted by apenwarrJosh Tabish @joshtabish.bsky.social · 28/09/2026“Canada should be able to support legitimate investigations without making secure systems easier to attack, and without making Canadian companies less competitive globally.” – Avery Pennarun, CEO of @tailscale.com in @betakit.com 182
apenwarr @apenwarr.ca · 29/09/2026An optimistic(?) data point here is that when unemployment spiked to ~20% during COVID, the US government (and you remember who was in charge) started mailing out cheques. GDP was also down, so this caused inflation to rise; but if GDP had been skyrocketing it would have been fine, sustainable UBI. 030
apenwarr @apenwarr.ca · 29/09/2026Oh yeah, there’s no guarantee that working *conditions* will get better, I’m just really skeptical that we’ll run out of work. That said, the history of tech has been working conditions improving for tech workers. And maybe we will all soon be tech workers… 040
apenwarr @apenwarr.ca · 29/09/2026Global poverty has dropped precipitously and purchasing power per capita has increased across the board 120
apenwarr @apenwarr.ca · 29/09/2026That’s an interesting theory but is the opposite of what I just said always happens. :) Even the greediest capitalist will hire more people when ROI goes up. 220
apenwarr @apenwarr.ca · 29/09/2026I think the U-6 unemployment rate measurement nowadays captures the cohort of discouraged and underemployed workers. U-6 unemployment in the US has been trending downward this year 120
apenwarr @apenwarr.ca · 29/09/2026Because ROI goes up when costs go down, causing companies to invest in expansion. Hiring plans are typically calculated in incremental revenue per incremental headcount. Incremental revenue accelerates when you do a bunch of high ROI projects 140
apenwarr @apenwarr.ca · 29/09/2026Were they put out of work because revenue went up or revenue went down? My understanding is the latter. Any those jobs are gone but current unemployment rate is very low 110
apenwarr @apenwarr.ca · 29/09/2026Sure but why do we think that will happen? Revenue went up so we fired the human workers? That doesn’t match any history I’ve heard. Also the economy stops working if buyers don’t have any income. Henry Ford was on the right track there. Capitalists are greedy, they know they need buyers to exist. 470
apenwarr @apenwarr.ca · 29/09/2026But there were immediately a lot of new factory jobs (which were presumably worse jobs I guess). I’m not sure how long the artisans were literally unemployed for. 220
apenwarr @apenwarr.ca · 29/09/2026The videoconferencing is clearly analog based on how it responds in low signal environments. This is pretty realistic for cross-species comms standards in my opinion :) 030
apenwarr @apenwarr.ca · 29/09/2026I mean, hmm, the industrial revolution? Surely that had a pretty wide scope. And I’m not sure what “too fast to adapt to” would look like. GDP and production/construction capacity so high that none of us can afford food or housing or useful things to do? 3130
apenwarr @apenwarr.ca · 29/09/2026I think this is the Malthusian argument paraphrased. Malthus doesn’t have a super successful prediction history 040
apenwarr @apenwarr.ca · 29/09/2026Arguably those are labour saving devices. Which did not reduce employment :) 150
apenwarr @apenwarr.ca · 29/09/2026I dunno, factories and printing presses were pretty unprecedented! And telecommunications and Wikipedia. It’s interesting to go from “never happened, not even once” to “will all happen in one shot.” 2420
apenwarr @apenwarr.ca · 29/09/2026We eliminated those menial jobs but created even more jobs to replace them. 131
apenwarr @apenwarr.ca · 29/09/2026I hear you, and raise you washing machines and electronic calculators 040
apenwarr @apenwarr.ca · 29/09/2026Jobs per person has remained flat, but number of persons has increased exponentially 2240
apenwarr @apenwarr.ca · 29/09/2026So far no labour saving device in the history of humanity has resulted in a long-term reduction in employment. 2920932
apenwarr @apenwarr.ca · 29/09/2026> It's a good book. You won't like it But worst of all you won't be able to prove me wrong unless you read it 230
apenwarr @apenwarr.ca · 29/09/2026I see endless debates about rationalists. Here's the book that completely disabused me of the notion that rationality (which I love) is a solution to anything. It's a good book. You won't like it www.amazon.com/dp/B005LXUSY...amazon.comOn EquilibriumAmazon.com: On Equilibrium eBook : Saul, John Ralston: Kindle Store 5233
apenwarr @apenwarr.ca · 29/09/2026Hmm. Rare situation where I think I’d like to be the product of choice in a *smaller* market segment please 0190
apenwarr @apenwarr.ca · 28/09/2026Repetitive Self Improvement Recursive Stress Improvement Recursive Stress Injury 3320
Reposted by apenwarrJanet 🇨🇦 @jansplanet.bsky.social · 28/09/2026“Canada should be able to support legitimate investigations without making secure systems easier to attack, and without making Canadian companies less competitive globally as more countries look for trusted, sovereign infrastructure,” #CdnPoli #BillC-22 #Security #Sovereignty betakit.comTailscale, Windscribe co-sign open letter asking feds to further amend Bill C-22 | BetaKitCanadian and European tech companies still have concerns over Canada's federal “lawful access” bill. 0249
Reposted by apenwarrAndree Toonk @atoonk.bsky.social · 28/09/2026This weekend’s project: DPDK + WireGuard 🤓 I wired wireguard-go up to my packetio library and got a single tunnel to ~30M pps clean, peaking 40M pps. That’s 70× what kernel WireGuard can do! Wrote it all up here if you’re into that kind of thing: toonk.io/wireguard-dp...toonk.ioWireGuard at 40 Mpps with DPDK, in GoOver the last few months, I’ve been spending a lot of time experimenting with high-performance packet processing in Go. A lot of that work turned into packetio, one Go API for sending and receiving pa... 0179
apenwarr @apenwarr.ca · 27/09/2026That’s all fine if everybody plays by the rules and agents don’t go rogue. The current legal landscape is rogue agents committing crimes and nobody’s being held responsible. So my joke-not-joke is the enforcement will get real once The Mouse is at risk 020