psimyn (@psimyn) profile picture

Stylesheet Computer

Melbourne, Australia
Joined April 2009
Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)Photo by psimyn (@psimyn)
slowing the fuck down still cool at the end of 2026.
I just finished talks about my software factories at two conferences. Afterwards a bunch of people came up and told me what they were surprised about. Here is a distillation of what people thought were hot takes from my two talks. First hot take: I tell everyone to slow down. Most speakers go on about how fast they're going, but I strongly believe in pacing AI outputs to human speeds. I regularly tell AI-pilled CIOs and CTOs to chill: almost nobody is _actually_ going fast with AI. Engineering still moves at engineering speeds, and business moves even slower. The ones pushing too hard are starting to see downtime for it. If you don't pace your AI to human speeds, everything starts breaking around you, even at small operations like my online video game. I only have about a dozen humans working with my factory, and only a few hundred active old-timer players in the community who get along well, and still, introducing AI has still been very rocky. My agents have, at times, overwhelmed my players with features, my accountant with emails, my merge queue with tickets, and my machines with load. Agents have pissed off my admins, resulting in name-calling, they've made messes in the game, and they've launched broken OSS without permission. The list goes on. I can get away with all this chaos because everyone involved in my operation is very forgiving. They're excited by the potential. But even so, I had to learn to slow down, real fast. And so that's what I preach. Hot take #2: Should you buy a software factory? Do you even need one? No. You probably don't need one. Factories take an extraordinary amount of upkeep. They turn into garbage mountains unless you put a frightening amount of time and tokens into curating them. I had to burn my previous five factories down since it eventually became cheaper to rewrite them than to fix them. So I'm not out there advocating for software factories or swarms just yet. If you do want one, be aware that it is a significant commitment of both money and human leadership to keep one running. Several people told me that I was the only one who stood up onstage, at either conference, and said Build, don't Buy. But my experience is that software factories grow to be highly bespoke to your domain. Mine is super tangled up with my game, and I think that's inevitable. Because the factory is not just building my product, it's supporting it too. It's a company in a box. And they're easy for Claude to build for you. Third and arguably hottest take: Next year will have a lot of human/agent and agent/agent office drama, resulting from bringing in AI employees, and it will be gloriously funny, but also frustrating. Many of you are already on this bandwagon, but everyone next year will have a personal chief-of-staff agent. Busy people, even hobbyists, will have whole entourages of helper agents. The hot take is that people will become so dependent on these little private entourages, that they'll bring them to work. Your people will talk to my people, etc. Whole protocols will emerge for private entourages negotiating work between humans, inside companies big and small. Nearly all humans will wind up having to start working with AI employees next year, and it'll be the beginning of the biggest cultural change we've seen in a hundred years. So that's the hot takes. Slow down. Only build a factory if you think you really need one. And focus on empowering yourself with your own agentic chief of staff (or team). Your entourage will define how you work with others, so start by perfecting how you work with them yourself.
18
21
5
602
46,623
The anti-AI backlash in games is mostly hurting indie studios. Insisting on no AI code in a world where all coding will effectively involve AI is an impossible standard. If fans care about human work, reward games centering human artists & designers rather than policing AI use.
The reveal that Big Walk used AI for "under 0.1%" of its code has caused a big blowup among fans, developers, and game award judges. Where do people draw the line on genAI in gaming, and how did a 2026 GOTY indie darling become ground zero for the debate? We broke down the messy discussion: kotaku.com/big-walk-ai-discl…
174
68
21
1,176
41,340
australia continues to build nothing, know nothing but try to control everything
In a randomised trial, software developers using AI took 19% longer to complete tasks. Yet they believed AI had made them 20% faster. At the ‘Evidence in Action’ event, I argued for a government that tests what works, and saves money by learning from the results. ministers.treasury.gov.au/mi… #randomistas
Readers added context they thought people might want to know
The 19%/20% figures are from a July 2025 METR study of experienced open-source developers using early-2025 AI tools, which the authors describe as an outdated snapshot; a 2026 follow-up found some evidence for speedup with newer tools. metr.org/blog/2025-07-1… metr.org/blog/2026-02-2… arxiv.org/abs/2507.09089
14
6
1
187
5,472
My $0.02 1) The rationalist/EA world has had far better intuitions about AI than the vast majority of the public for a very long time. 2) Anyone who's been practicing their argument for years WILL beat you in an argument, but that's distinct from them actually being right.
Replying to @loubie @TheStalwart
Personally, as someone who works in STEM but is outside of these ideological circles. The one thing i always found unappealing with Rat / EA thinking was the sheer overconfidence in “out driving their headlights” in terms of predictions. Extrapolating dozens, if not centuries, into the future with their maths just to win these hypothetical internet arguments. There is just never going to be enough evidence truly make these predictions anything more than vibes. Which is why even among the doomers the variability is essentially 2-99%
40
17
8
849
73,392
We’ve known since 2023 AI could not continue to improve due to “model collapse”—training on poisoned internet data generated by AI. This is why LLMs have made no progress since that time, steadily declining on all benchmarks, and are projected to disappear completely by 2030.
135
103
25
3,257
220,472
psimyn retweeted
this is one of those moments that makes you wonder what kind of information you've been ingesting from areas you aren't familiar with
“Is this just a big circle jerk?” Jon welcomes Ed Zitron (@edzitron), host of the podcast Better Offline, to make sense of the financial house of cards propping up AI and big tech. New pod out now!
63
47
11
1,473
107,890
psimyn retweeted
every launch vid now is like *cavemen making fire* *model t* *wright bros* *moon landing* *nuclear bomb detonation* “every great innovation was once called crazy… that’s why today we’re launching AI-powered automated accounts receivable for operationally complex services businesses”
52
155
24
2,744
103,270
"University researchers built a satellite that tracks people from low Earth orbit and dispatches a 5 kg payload to within .1 meter of their position, traveling at Mach 25 to anywhere on Earth in under 5 minutes. They suggest it could be used to deliver greeting cards."
Purdue University's IDEAS Lab built an autonomous drone that tracks runners through dense forests completely without GPS or cloud networks. By running 3D LiDAR mapping and obstacle-avoidance AI entirely on-chip, it can weave under tree canopies and predict target movement through blind spots in real time. Massive implications for next-gen search & rescue and edge AI.
370
3,114
153
42,660
629,763
Replying to @valvesxyz
I like to look at claims like this in case I can improve Ghostty. I grabbed your source and immediately spotted the issue: your read loop overwrites a 64 KiB ring. Everything older is discarded unparsed per tick. You're faster because you're not actually terminal emulating. It means the number isn't a throughput number. In a kitten benchmark run at the rate you're claiming, roughly 6.6 MB arrives per 60 Hz tick and 64 KiB of it gets parsed. You're read(2)-ing 100% of the data and parsing about 1% of it. There are other minor issues (a small fixed scrollback in a preallocated ring buffer, no grapheme clustering, throwing away data like OSC 8 URIs, etc.), but the major one is simply that you're discarding ~99% of the input. For plain text, e.g. your cat benchmark, the final screen ends up looking right, but the scrollback is missing 99% of the file and each retained chunk starts mid-line. As soon as there are escape sequences in the dropped spans, your emulator is fully out of sync and invalid: a cut inside a CSI prints the parameters as text, and any SGR, mode, title, or alt-screen change in the gap is just... gone. FYI, this is why I don't like when people point at refterm as a terminal benchmark. refterm implements a TINY subset of what a terminal emulator must support, it's closer to a fast file reader, not a terminal emulator. I also don't mean any shade by this. I like to look at technology objectively to try to learn something and improve my own craft and work. So I'm sharing what I found. I modified your code very slightly to parse on every full buffer (rather than per tick). And the results are attached. Both use main commit in each repo at the time of this post. Note the graph is in log scale, the discrepancy between performance for things like Kitty images is measured in GB/s.
180
242
215
8,325
866,979
How does this nonsense make it into the paper of record?
140
162
66
3,832
413,371
One thing about abliterated models is that it makes you very impatient with the non-abliterated models’ refusals to do research that they deem “unethical” for basically subjective (often bad) reasons.
12
24
1
316
13,997
went to visit a friend who’s a ceramicist and works in a studio with no phone signal in a tiny yorkshire village. he doesn’t listen to podcasts or music, only occasionally uses the radio, and when I asked him what he thinks about all day he just said “well, pots mostly”
257
2,327
665
43,951
952,276
If you Moneyball your startup idea, you end up selling data If you follow your emotional startup idea, you end up building a personal assistant If you build something that people want, you end up forward deploying AI into car dealerships in middle America
46
51
12
1,499
68,246
psimyn retweeted
I've been cooking on this post for a bit, but with the cyber security events of this week I had to ship it: Long-live static credentials must die. We never want to rotate API keys again. Turns out only 8% of cloud and SaaS vendors support secure alternative. This is our call to action to change this AND our commitment to do our part: - We are requiring vendors on our marketplace to support OIDC - We shipped @shadcn components to make it easier to ship OIDC configuration - We are expanding OIDC support on our own platform The last generation got the web from HTTP to HTTPS. It is on us to move past API keys. vercel.com/blog/death-to-lon…
70
93
30
958
72,349
Railway is compromised, rotate all API keys as soon as possible Deployments are down so don’t wait to redeploy Attackers are draining keys in real time, we didn’t lose much but can confirm this is still an active attack Stay safe friends
121
490
173
4,723
568,445
Just going to preregister here that I believe the computer security overhang is not a few dozen bugs that get patched and then we're good, there are 100,000s of vulns that are extremely discoverable with infinite free tokens and we're probably going to have to move away from systems that presuppose that perfection is possible.
Fascinating stuff. I'm not an expert, but I remember that dodgy CAs have been a known problem since at least 2010, data retention/version control leaks at least as long, and PHP has always been adjacent to self harm. It'll be interesting to see where we land in 12-18 months on this stuff.
37
29
8
602
35,472
I sometimes want to give up on progressive web apps. The obstruction by Apple, companies that only use PWAs to promote their native apps, fragmented browser support, developer ignorance, and half-baked examples that enforce the idea that web apps are inferior. It feels like an uphill battle that never ends, and I sometimes lose motivation to fight it. But what if this fight is not fought? We end up without this free distribution channel called the web, and apps will only be created inside walled gardens called app stores where Apple and Google decide the rules. Where they can extract heavy taxes and decide who gets in and who doesn't. Sure, we can still put our apps on the web, but as long as Apple is able to kneecap web apps in such a way that there's no level playing field, will they ever be viable? That is still a fight worth fighting, and I will continue to show what web apps are capable of with What PWA Can Do Today. That is my motivation. Even if certain things only work on Android or desktop. Even if Apple makes installing a PWA more cumbersome with every iOS release. Thankfully, we have people like the Open Web Advocacy fighting the legal battles. We, as developers, can and must show the world what web apps are capable of. Build great freaking web apps. Even if some features are only available in Chrome or on Android or desktop. Otherwise, the world will never know what web apps can do. The free, open web is at stake.
8
13
4
113
4,094
psimyn retweeted
we're entering another wave of attempting to make the web more interoperable but i am not sure if anything important has changed idea is simple - if apps published APIs in a standard spec this would allow for maximum interoperability pretty much the only system we use day to day that works like this is email - can use it with any client/server and talk to anyone even if they made different choices it's a nice ideal: - would be easier to embed posts from other sites - could have alternate clients to popular networks - could trigger actions on other platforms from anywhere there are two dimensions to this problem, technical and economic the technical solutions historically sucked - trying to model every possible case is effectively impossible saw this with semantic web and smart contracts in crypto but LLMs seem to solve this - you actually don't need an intense standard anymore they can figure out how to call any system so that leaves the economic which is the bigger issue there really is very little incentive for interoperability - no companies want to give up the last mile uber eats has a very specific checkout flow that is optimized to get you to spend more why would they publish an MCP server that releases control of that? it would only happen if there was overwhelming consumer demand like with mobile - if you refused to publish a mobile app you effectively died and even if there is, they might just do one off integrations with chatgpt / claude as much as i want the ideal to happen the underlying economics push things towards centralization and walled gardens
40
20
15
742
91,511
lift ten times your weight. voraciously consume sugary liquids. kill anyone who doesnt smell like you. ant mode.
146
20,718
344
222,183
3,124,454
psimyn retweeted
Something you realize if you pay attention to travel time is a lot of fast driving is actually pointless. It’s rushing to stop lights.
This was my experience in Waymo. They don't edge into crosswalks. They don't race around stopped cars. They're not aggro with cyclists. But also, they're (somewhat surprisingly) not slow. They simply follow the rules to the letter.
37
35
12
1,749
33,874