I observed the same thing happening when we transitioned from analog land line telephones to everyone having mobile phones. The plain old telephone was crispy clean, and cellphone calls by comparison were just barely intelligible. It's so bad that I'm honestly surprised that texting didn't take over sooner. It wasn't until Apple FaceTime Audio that calls started to sound good again, and it still isn't the default.
Multiplexers have a lot of little papercuts and limitations when you want consistent and native-to-the-OS behavior to flow all the way between the OS and the apps and connections in the multiplexers. For example, gesture scrolling in a tmux/zellij pane is hard because the right events don't flow through. It's an itch to scratch like designing a good terminal lib (like ghostty.)
And then, there's all the unifying and smoothing out of data connections--I imagine they'll be a lot to make sure that, e.g., an ssh session or a Claude code session in a pane can be seamlessly reconnected or never drop when the pane is closed. This takes a lot of hacking to make look seamless in most setups right now.
Like mitchell, I write a lot of software around multiplexers, and I've thought about making something similar, but I don't have quite the same level of personal wealth or angel investor connections. :)
herdr and mosh are a couple of reference points for this, there are others.
Yes, but you can't, for example, transfer momentum vectors through zellij to the app.
If the app is offering what appears to be momentum, it's implemented its own acceleration/deceleration algorithm. That's a lot to ask of an app, and they aren't equally good.
Detecting a wide range of underlying terminal support for a CLI app is a pain and prone to errors/drift.
There are various plugins that attempt to solve this, but it'd be nice for someone to solve it once and for all, and then upstream those layers into open source as this project is promising to do.
I've wondered this myself. I stuck with iTerm2 for a really long time but recently switched to Ghostty. Ghostty definitely seems to renders faster (I say "seems to" because I didn't benchmark or anything).
I don't actually use a multiplexer. I'm not sure I understand the use case when most terminal emulators supports tabs + windows (like iTerm2 does). Or you could just launch multiple windows and use your OS's window manager to "multiplex". Are the terminals connected in some way that I'm missing?
Regardless, I think Ghostty is a reminder that you can win in a space that's crowded. There's probably a lot of widely used software that would get dropped en masse immediately if someone had the guts to make a better alternative.
Typically, the benefit of the mux over just more windows or tabs is the ability to detach sessions from a term and have them continue to run. But yea, the “split a term window into multiple terms” doesn’t have quite as much value in a world of lots of tabs. That said, you can do it under program control on the remote host, so it does save remote login overhead, etc. But for local, less value.
I don't get it either. What's wrong with Terminal and tmux, screen or what have you?
I don't get the whole GPU rendering thing for ghostty either. Its a terminal emulator. It shows plain text. Why would I need GPU acceleration? I never felt "yeah, this terminal needs faster rendering".
If you like terminals and tmux, can I suggest you spend a few hours playing with herdr? It clicked for me in a way that a lot of other orchestrators didn't.
Company scrip was company currency and employees were required to be paid in it and could only be used in company stores. Employee stock is getting some of your pay, voluntarily, in stock which can be sold in market. As long as company doesn't own housing or run grocery store, it isn't scrip.
At this point it might be easier to point out articles not written by LLM. Most of them are, I just can't bring myself to read them anymore, it feels like a waste of time. I want to read original thoughts and it's almost impossible to know if this kind of content contains any. I'm tired.
I actually find it hard to read. Something about it is hard to follow, and I have to think very hard to keep the ‘story’ straight in my head. I keep having to re-read sections, which it not normal for me.
This doesn’t make much intuitive sense to me because shouldn’t output based on statistics have a very strong through line?
This is where I really first got lost:
“You can’t run a greenhouse without gas unless you can see the moments it slips — a fault, a silent fallback — and close them one at a time. The monitoring doesn’t undermine the gas-free promise. It’s what makes the promise keepable.”
Possibly a case of a non-native English speaker using AI to translate or help them do their write up. In either case, it sadly makes the reading experience less enjoyable.
Is laser printing more feasible? I've had the same Brother black & white laser for almost a decade and it's more than adequate for everything I do. If I ever need color, I send it off to a service that has much nicer printers than most people would ever want in their homes.
> If you encountered a cheetah in pre-industrial times (and survived the meeting), you might think it was impossible for anything to go faster.
Fun fact, there is no historical evidence of an adult human ever dying from a cheetah attack. They are naturally shy, and a lot smaller than you may realize.
> The Man versus Horse Marathon is an annual race over 21 miles (34 km), where runners compete against riders on horseback through a mix of road, trail and mountainous terrain. The race, which is a shorter distance than an official marathon road race, takes place in the Welsh town of Llanwrtyd Wells every June.
> ...
> The event started in 1980, when local landlord Gordon Green overheard a discussion between two men in his pub, the Neuadd Arms. One man suggested that over a significant distance across country, man was equal to any horse. Green decided that the challenge should be tested in full public view, and organised the first event.
While the horses had a string of wins from 2008 to 2019, 2022 to 2025 had three wins for humans and one win for a horse.
The Man vs Horse marathon is interesting but in a real race we have no actual chance of winning.
There are horse endurance races where the winner arrived in 7,5 hours after 160km[1]. That's a sub 2-hours marathon almost 4 times in a row (not to mention with a guy on your back).
“in a real race we have no actual chance of winning” is an absolutely wild thing to say in response to a link to a real race in which the human has won the last few years in a row.
The race is pretty much designed to be a difficult for horses as possible to give humans a chance. Except for the parts that are extremely difficult for horses, horses steamroll the human competitors.
Wow, seems surprisingly balanced. I would guess that if it was 50% longer distance than humans would win reliably, and 50% shorter would allow the horses to win?
Think it's more to do with hilly, wooded courses unfavourable to horses and a requirement that the horse's heart rate remain below a certain threshold when inspected by a vet in the middle and end of a course, whereas the leading humans are maxing out their efforts
I don't think the Mongol cavalry would lose races to humans over any distance of steppe
> Lobb’s victory came on a hot day, as did Florian Holzinger’s subsequent victory in 2007—a significant detail, according to a new study in the journal Experimental Physiology from Lewis Halsey of the University of Roehampton in Britain and Caleb Bryce of the Botswana Predator Conservation Trust. Halsey and Bryce gathered historical data from three endurance races that pit humans against horses, including the Man Versus Horse Marathon, to test the idea that humans are uniquely adapted to run for long distances in hot weather.
> This idea has been around since the 1980s, and it gained prominence when Harvard anthropologist Daniel Lieberman and University of Utah biologist Dennis Bramble published a 2004 Nature paper hypothesizing that running had “substantially shaped human evolution.” They argued that our ability to keep running at a moderate pace even on hot days allowed us to run prey like kudu to exhaustion or outcompete other animals in the race to scavenge carcasses left by other large predators.
There's a plot with the analysis of the Old Dominion with weather stations in there that show a steeper negative slope for horses compared to the humans.
> Overall, for every increase of 1 degree Celsius (1.8 degrees Fahrenheit), the horses slowed down by about 1 percent—or 0.07 miles per hour, to be precise. The humans, on the other hand, slowed down by just 0.04 miles per hour for each extra degree of heat. That 36 percent advantage for the humans was statistically significant.
---
For the Man vs Horse, the weather conditions ("Hot", "Rain/sun/windy" - not exact values)... the entry for 2022 was the human winning by 1:51 on a warm day, and 2023 was a human wining by 9:44 on a sweltering day.
> ... There are other Man versus Horse races — in Scotland based at Dores, near Loch Ness, in Central North Island, New Zealand and in the U.S. city of Prescott, Arizona.
I think the term is persistence hunting (https://en.wikipedia.org/wiki/Persistence_hunting). There was an intriguing blog post years ago (which sadly I can’t find) about how terrifying a fictional portrayal of persistence hunting would be- grinding down prey through exhaustion, thinking you’ve escaped but being slowly tracked down.
To what end? Such that you can wait out your days in a secluded bunker posting online trying to convince anyone to believe you as to the reason you've put yourself in the bunker?
Because otherwise, 'it' will just back out of your trap and go along to continue to follow, wouldn't it?
Fun fact for me personally: because I am lucky enough to know a zookeeper who was in charge of the ambassador animal program at a major zoo, I have direct personal experience that cheetahs purr when you pet them.
You're right. They're smaller than you probably imagine (about the weight of an average Labrador). That's still definitely big enough to be a problem if they felt threatened, I'm sure, but the animals my friend was in charge of were raised to be around people for outreach purposes. That particular cheetah, for example, had once been on the Today show.
Cheetahs (genus “Acinonyx”) are the largest ‘small cat’, apparently [1] - separate from the genus “Panthera“ containing the ones that will attack humans (lions, tigers, leopards).
[1] There’s a well known viral video of a wildlife park keeper who sleeps with three cheetahs who behave pretty much like large house cats.
Yep. That said, unlike cheetahs, there’s plenty of evidence of leopards attacking humans. And these days, it’s the leopards, the closed-AI types and misanthropes -- telling everyone, “AI will take your job and there’s nothing you can do about it.”
When Dario and others say things like "this is happening and we should probably figure out what to do about it" what ends up happening is people hear "this is happening," see that the person warning them is the person doing the thing, and then short-circuit. "Why can't you just stop then?"
Dario's point, and the point of the people actually trying to solve the problem, is that AI is not just Anthropic and OpenAI. It's the knowledge that you can put more compute in, and get more capability out.
It is a technology now. It exists, in the world. Wishing will not make it go away. Being angry at it will not make it go away. Lying about how much water it uses will not make it go away. If Anthropic and OpenAI Shut down tomorrow, Accenture will not say "oh guess that llm thing won't work, let's go back to hiring humans!"
It is a truth that you can multiply matrices and get something that is economically useful. We cannot un-know this.
Physics allows it, so it will happen. So we should probably figure out what the heck to do about it. If your answer is something along the lines of "restrict it" then 1. let me know how that goes when other people don't, and 2. I really would rather prefer a world where we have the machines do the work the machines can do, not a world where we have human makework. If this means we need to figure out redistribution, let's talk about redistribution!
Physics allows this, and actually taking advantage of it requires billions of dollars of unprecedented infrastructure buildout that is already destabilizing the power grid.
The only reason that infrastructure buildout is happening at all is the ideological capture of a small handful of obscenely wealthy people, who are fueling this buildout by spreading the extreme paranoia you’re echoing here.
I do not understand why no one else can see the circularity of this reasoning. There is nothing inevitable about tying up all of this productive capital in the pursuit of AI. There are many, many other projects requiring similar capital and human effort, with much more obvious payoffs, such as decarbonizing the world’s energy systems.
“It’s physically possible to provide abundant electricity without burning fossil fuels” is more provably true than any of the insane science fiction bullshit that undergirds the AI buildout, and yet, the entire clean energy industry is still having to build insane financial Rube Goldberg contraptions to make incremental progress.
“Inevitability” is a lie, period. This entire thing is extremely historically contingent, and we could easily stop this train tomorrow.
> This entire thing is extremely historically contingent, and we could easily stop this train tomorrow.
So, the Baruch Plan?
The Manhattan Project was $~2B in 1945 dollars, and a national-scale industrial mobilization. Now North Korea has the bomb. That's with nuclear material, which doesn't get easier and easier and easier to work with every year.
Compare to the price to train GPT-2 in 2019 ($43,000), and in 2026 ($73) [0].
Since nobody uses GPT-2 any more it's more informative to compare the price to train GPT-2 in 2019 and the price to train GPT 5.5 in 2026. Unfortunately that cost is not disclosed but it's probably in the billions.
The point being: the price to train frontier models isn't coming down, nor is it going to come down because for models to remain on the frontier they have to keep getting bigger and bigger (and trained on more and more data).
> There is nothing inevitable about tying up all of this productive capital in the pursuit of AI.
In the US capitalist context, it's certainly inevitable, because AI is the biggest and most attractive source of profit and power out there right now. In that context, the broad strokes of what's happening currently, including the financial bubble, are predictable and inevitable.
What are the concrete steps which would allow us to "easily stop this train"? And why haven't we used steps like that to stop other cases where obscenely wealthy people have screwed everyone else over to increase their wealth? Is public control of the means of production involved, perhaps? If so, your definition of "easily" and mine are incompatible.
I grew up in a union household, and my dad and my grandfather fighting for better wages, healthcare and working conditions are the reason why I got a good education and work in Silicon Valley surrounded by Stanford assholes.
All of us who work for a paycheck can get together and say, “no, we will not allow you to record keystrokes and mouse movements to train our replacements. No, we will not have our performance or future employment based on an AI leaderboard.”
Previous generations fought and died for our right to do that, but in 2026 we just sit on our hands and complain on this forum. We can and should do better.
The U.S. is absolutely on fire right now with opposition to data centers. We, collectively, can extract concessions or ban their construction altogether.
These things aren’t “easy”. They are also eminently possible.
I think you're in denial about the reality of the situation.
The reality is that the US has been a story of increasing concentration of wealth and power. The people who "fought and died" bought some important (from a human rights perspective) but ultimately minor (from a capitalist perspective) concessions from the capital class. The battle you describe is one of defense of rights, not of gaining control.
The overall system of capitalist control remains unaffected, and it's why the buildout of AI is, in fact, inevitable under the current system.
You're essentially saying no, you want public control of the means of production instead. That might be great, if even a sizable fraction of the US population agreed with you. But due in no small part to decades of propaganda, they don't.
A US-China AGI ban treaty could prevent superintelligence indefinitely. Data centers are hard to hide. Have fun buying GPUs when you're cut off from all global payments. America would have to make some unpleasant concessions but that seems like a solid trade for preventing a wide variety of nightmare futures.
I don't think we can stop it, but the people saying "this is happening and we should probably figure out what to do about it" are doing their best to accelerate it, which gives us less time to figure out what to do about it.
The people saying this is happening are competing with each other for resources, so there's no way for one of them to hold back without losing out to the others. We see this with people dropping Claude subscriptions in favor of ChatGPT because codex 5.5 > Claude Opus 4.6/7/8. Anthropic is losing money by not releasing Mythos.
Aidenn0 said they're doing their best to accelerate it, and I'm saying "they're" isn't a single monolithic entity and that they're in competition with each other so they're incentivized to go as fast as possible, so it would be hard to hold them back.
> Aidenn0 said they're doing their best to accelerate it, and I'm saying "they're" isn't a single monolithic entity and that they're in competition with each other so they're incentivized to go as fast as possible, so it would be hard to hold them back.
Maybe it would run afoul of antitrust regulations, but it's totally realistic for all those competitors to get together and say "hey, we could really fuck up society in our race to get rich with this tech, lets all slow down." And if these companies are run my mature people who don't subordinate every consideration to greed, they'd to it.
I would propose that it is very likely to have zero effect. Your argument supposes they are all working together, like many connected computers calculating primes.
It only takes one of them to do it and they are not sharing information. If the 1 you remove from N is the one that will discover it, then it will dramatically affect when AGI happens. If it is not, then it will have zero effect.
> When Dario and others say things like "this is happening and we should probably figure out what to do about it" what ends up happening is people hear "this is happening," see that the person warning them is the person doing the thing, and then short-circuit. "Why can't you just stop then?"
I still believe Dario asks these questions in good faith. Nobody believes that about e.g. Sam Altman or Elon Musk. They compared themselves to Oppenheimer because it helped them get attention. When it started an actual regulatory conversation, they were suddenly less worried.
