Jake Gold @jacob.gold · 11hThere were maybe two good moments in the whole movie, which isn't nothing, but it's not much. 010
Jake Gold @jacob.gold · 11hIn the bad ol' days, I watched a team rewrite a Perl codebase in Java. After a year of waiting, the CEO was surprised to find that it was missing a bunch of features and had no other perceivable benefit to him. They went back to the Perl codebase and feature development resumed. 000
Jake Gold @jacob.gold · 12hMost unbelievable scene in Disclosure Day wasn't the aliens. It was the: 1. Over the shoulder bread holder 2. Stealing a chunk of plain bread 3. Insufficient toasting 4. Dedicated butter dish 5. Butter dish stacked on bread Spielberg is an alien that doesn't understand toast. 2191
Jake Gold @jacob.gold · 17hIf Google is still Google then "algorithms and datacenters" is definitely their game to lose. But maybe the brain drain and culture rot has gone too deep. 2250
Jake Gold @jacob.gold · 18hExactly, so calling it a three horse race ahead of anyone even using it is being a silly goose. 220
Jake Gold @jacob.gold · 18hYeah, I'd imagine that felt somewhat like NASA and other government agencies. 010
Jake Gold @jacob.gold · 18hFrom what I've seen, most <strongly methodological> projects turn into non-<strongly methodological> projects halfway through. Or they fail. At some point the productive devs on the team say "do you want me to do it your way or ship the damn software?" and then managers make the obvious choice. 240
Jake Gold @jacob.gold · 19hReally enjoyed The Odyssey But surprised and a little disappointed to say that the Ralph Fiennes movie The Return did the final scene much better. www.imdb.com/title/tt1986...imdb.com 050
Jake Gold @jacob.gold · 19hIt's weird that Slack hasn't added support for ACP + diff viewer etc. So that using agents via Slack wasn't such a huge downgrade vs using them with a purpose-built interface. Slack is already fine for non-devs but I don't see how devs live with generating code, it would slow me down so much.agentclientprotocol.comIntroduction - Agent Client ProtocolGet started with the Agent Client Protocol. 000
Jake Gold @jacob.gold · 20hYeah I believe you that people did it. NASA was doing something like waterfall in the 1960s from what I understand. But the Macintosh team wasn't doing waterfall in 1983. 130
Jake Gold @jacob.gold · 21hI didn't start until 2002 but pretty sure it always a lot on depended on the company. 130
Jake Gold @jacob.gold · 21hIn practice most teams, from what I saw and know, would give up half way through and have some small number of devs save the project by bypassing all the formal processes. So yeah lots of people tried bad things that didn't work, full waterfall, full Agile, etc. None of it worked. 130
Jake Gold @jacob.gold · 21hSoftware development improved and agile contributed to that, but so did rapid prototyping and cowboys and the rest. Everyone gave up on highly formal processes. What works is hiring great people and shipping incrementally. Software moving to the cloud had more to do with this than any methodology. 170
Jake Gold @jacob.gold · 23hWhat? Agile did not speed up software development by 10x. With respect, that’s just a silly claim to make and there’s no data that supports it. People actually following Agile (which almost no one did) probably moved slower. Open source libraries sped things up but probably not even by 10x. 4420
Jake Gold @jacob.gold · 30/09/2026About Astra? Bigger models are just generally much better at avoiding prompt injection attacks for mostly same reasons they’re better at everything. 010
Jake Gold @jacob.gold · 30/09/2026San Francisco Startup Rule #5823: you're not allowed to judge a person for ditching out on an important meeting if they're going to see The Odyssey in IMAX 70mm the day before it stops playing at The Metreon. 1140
Jake Gold @jacob.gold · 29/09/2026I’m optimistic that some of this is true but it’d be extremely lucky if it was fully symmetric. If we experience any algorithm/architecture breakthroughs we could quickly end up with order of magnitude improvements, equivalent to building the hw 10-1000x faster. Will be exciting to find out. 100
Jake Gold @jacob.gold · 29/09/2026“What if we flip the b around?” Smart using Astra so that it’s resistant to prompt injection. 1110
Jake Gold @jacob.gold · 29/09/2026Okay Jev had less than two months in the sun. I should’ve predicted weeks. 1220
Jake Gold @jacob.gold · 29/09/2026I knew there was a reason I had an extra cup today, I just didn’t know why. Until now. Turns out I was celebrating. 1170
Jake Gold @jacob.gold · 29/09/2026Still, factory automation was one of maybe a dozen core technologies that made up the Industrial Revolution. AI sits on top of a bunch of technologies too but we’ve already built them. The industrial revolution was slowed down in ways, and to a degree, we can’t reasonably hope for with AI. 100
Jake Gold @jacob.gold · 29/09/2026If AI is as big as the Industrial Revolution that’s pretty scary The Industrial Revolution took hundreds of years to unfold, giving civilization time to adapt. AI will absolutely not take hundreds of years. We’re all just hoping it takes at least decades. 3110
Jake Gold @jacob.gold · 29/09/2026Seems worth dwelling on the idea of comparing the Industrial Revolution to AI, a specific technology. 110
Jake Gold @jacob.gold · 29/09/2026Starting to think this guy isn't a reliable source of unbiased information. 061
Jake Gold @jacob.gold · 29/09/2026We've had unprecedented technologies before, but each was relatively narrow. AI has the potential to replace a much broader range of tasks across many jobs. So "never happened" just doesn't seem predictive in this case. And it doesn't have to happen all at once, just too fast for us to adapt to. 180
Jake Gold @jacob.gold · 29/09/2026Sure, but the relevant history is extremely short and there are strong arguments that AI is unprecedented. So far it's just "more computer" but it's moving quickly enough that we just can't be confident one way or the other. 2260
Jake Gold @jacob.gold · 29/09/2026I've passed it like a thousand times just because it looks like it's not a well run location. I'm sure it is fine like you say. 010
Jake Gold @jacob.gold · 28/09/2026This is great. PLC being centralized is a legit weakness in the Atmosphere's story and this goes a long way to mitigating it. I still want @dholms.at to be the PLC CEO, regardless of whether it makes any sense or not. We had a deal, kinda. 1330
Jake Gold @jacob.gold · 28/09/2026"nothing more permanent than a temporary solution" personified as a Swiss legal entity 0360
Jake Gold @jacob.gold · 28/09/2026What we have today is stuff we can deal with, however bad it might be. The entire AI x-risk concern is about truly superhuman AI that no human expert can deal with because it's much stronger, faster, and better than anyone. Merely being skeptical of progress doesn't answer the hard question. 100
Jake Gold @jacob.gold · 28/09/2026It's never perfect and still makes pronunciation mistakes, but it's reached a level of quality that makes it very useful. It's not as good as a human tutor in that way but the endless patience and ability to use it in the shower and in car makes it life-changing for me. 110
Jake Gold @jacob.gold · 28/09/2026FYI, ChatGPT Voice mode is finally good enough to tutor you in Japanese! (and probably other languages) I've tried every version since it was released and it didn't work well enough. It's still not perfect, but the endless patience and lack of embarrassment make it incredibly productive. 2171
Jake Gold @jacob.gold · 28/09/2026Yeah, most people are talking shorter timelines than 50 years, but timelines for technology progress are notoriously difficult to estimate. Maybe it's 75 years, but maybe it's 5 or 15 years? It's just N breakthroughs away and no one knows, but we can be pretty sure it'll happen. So then what? 120
Jake Gold @jacob.gold · 28/09/2026Aliens would almost certainly be so much more advanced it wouldn't be worth worrying about. Do you actually doubt that we can re-create the functionality of a human brain in a computer, within say 50 years? 200
Jake Gold @jacob.gold · 28/09/2026I feel exactly the same way about the kind of LLM-based AI we have today. But the AI x-risk argument is about what happens when we have "real" AI that can actually think like humans but a million times faster. There are no real experts on this and yet it does seem worth worrying about in advance. 170
Jake Gold @jacob.gold · 28/09/2026I'm not sure how valuable his ideas were, but we can't argue that they weren't influential. And yeah, definitely not comparing Darwin's discovery to Yudkowsky's ideas by any means. 000
Jake Gold @jacob.gold · 27/09/2026There’s no amount of cope on my end. I use and enjoy this stuff all day long and know how powerful and flawed it is. I object on technical grounds to the misuse of words that mislead less technical people. If we had an “FDA for AI” they might rightly force companies to be very careful with words. 110
Jake Gold @jacob.gold · 27/09/2026I remember now that I already complained about this a year ago and I suppose I should keep up the tradition going forward. I should have "Grok Bot" remind me 🫠 030