Rendered at 11:45:36 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
themgt 18 minutes ago [-]
Gas Town was intended to be reusable, but I only ever wound up using it to build itself. Gas Town fell apart at the seams with Opus 4.7. Up through 4.6 it was working brilliantly. With 4.7 we saw the introduction of the "just two more things" tic, which prevented Opus from ever converging on being ready to do real work—it always wanted to fiddle with Gas Town itself. The Opus tic never went away, so Gas Town effectively burned down. It had other problems, too, but 4.7 was the final straw.
"True Gas Town has never been tried" etc.
With Wheelhouse, I have reinvented something strangely Gas Town shaped from first principles, but it's running many more agents, and they are far more organized. This has shown me the shape of the changes coming next year. I'm operating about 12 months in the future. I managed this only through luck: for the past 18 months I have had the magic trio of time, money, and energy, and I've been spending it all on figuring this stuff out. And I am grateful for the privilege.
I'm still half-convinced this is Yegge's version of Andy Kaufman's professional wrestling career, "Gas Wheeltown was an 18 month long sociological experiment conducted by Thiel University. The study is now complete, thank you for your time."
xnorswap 3 hours ago [-]
I don't like using the term "AI psychosis" because it originally meant a narrow definition resulting in genuine psychosis and acute mental health trauma, but seems to have widened to anyone falling down a much wider trapdoor of AI engagement.
But I read, or at least attempt to read, something like this, and I can't help but marvel at the capture that AI agents seem to have on some individuals.
I expected Yegge to, at some point, assuming this isn't still all satire, to take a step back and take stock that the basic harness are improving faster than the rate at which these complex beasts can adjust.
But the opposite has happened, he's dived fully into believing that the secret to unlocking the power of AI is to wrap more layers of AI around it.
If you actually want a problem worked on all night, just fire up fable, type /goal and then describe the goal. With permissive run settings it'll crunch for as long as it needs.
I've not seen any evidence that wrapping that in a further 3 or 4 layers of agents improves anything.
cmpxchg8b 2 hours ago [-]
I think the domain "yegge.ai" says it all. It's becoming like the NFT craze, and I say that as someone who uses AI every day.
inigyou 1 hours ago [-]
Ironically, AI tokens are fungible.
54 minutes ago [-]
deadlast2 16 minutes ago [-]
MOE and COT or github co-pilot. I think agents can marginally improve the performance of AI.
ramoz 2 hours ago [-]
I mean, the dude uses a lot of jargon, ego, and mythologizes stuff but the article comes down to a pretty simple practice:
Run a big task queue, have a smart model manage a cheap model on every ticket, set up a cron for models to automatically work off of events, stop gating merges on builds, and expect to spend a quarter of your time maintaining the machinery (all of it currently subsidized by many $200 accounts).
alansaber 27 minutes ago [-]
Yep. Though hard to write a compelling blog by being to-the-point without data
nautilus12 23 minutes ago [-]
I call it "yegge sweats"
xnorswap 3 hours ago [-]
I'm still on the fence whether Yegge is being satirical, some phrases are just too on-the-nose:
> Gas Town was intended to be reusable, but I only ever wound up using it to build itself.
petesergeant 1 hours ago [-]
I think he's both honest and well-intentioned without being self-aware
bambax 2 hours ago [-]
I used to like Steve's posts. They were all way too long, but a fun and interesting read nonetheless. He started to lose me when at Sourcegraph where he oversell Cody which didn't do anything useful (I even had a commercial meeting with Sourcegraph people who weren't of much help); then he completely lost me with his GasTown business where AI meets crypto (what could possibly go wrong with that). Now I can't read anything from him anymore ;-(
tclancy 1 hours ago [-]
It is a rough go for a long-time reader, the bragging tone of six figure engagements and “whales” (i.e., suckers) keeping the game going. The statement that what he is doing is what the reader will soon catch up to feels like someone who can no longer see land.
That said, as another nut with an agent factory toiling away around me, I’ve also started to think about replacing any kind of CI in the cloud with local. I couldn’t quite keep myself at the article to read what comes next, so I guess I will catch up to it down the road.
jwr 23 minutes ago [-]
I hear so much about people running into token limits and using bajillions of tokens. Am I doing something wrong? I'm finding it hard to run into the limits of a Claude Max (20x) account. Well, perhaps the 50% Fable limit is something I have to be aware of, but otherwise this more than I can consume. I am the limiting factor: my capability to review code and design documents.
I use it to develop a large MRP-style app. This is as complex as things get.
Is it Clojure that makes things so efficient? My work style (lots of design documents)? I really don't know.
Do people skip the "review the code and read the documents" step perhaps?
pringk02 2 hours ago [-]
> But my Wyvern development has been burning the equivalent of $87k/month of API token burn, or about 69 billion tokens in July (96% cache hits, fortunately). So I do have to worry about it. I don't care how much money you have; your spouse is not going to let you spend almost ninety thousand dollars a month on your hobby game. I didn't even bother asking.
> My solution has been to create a token tap on $200 Max accounts, which for me work out to ~30x the list-price equivalent. So in reality I'm only spending about $2800/month out of pocket for my $87k "worth" of tokens. Though that number keeps growing alarmingly.
But how much does the game bring in?
mad0 2 hours ago [-]
> Once you're hooked on Wyvern, you're hooked for life. I have whales that have spent thousands on it.
At minimum 4k (given by whales = min 2 whales, thousands = min 2 thosands :) ) So I'm guessing that this covers Claud accounts.
"$5K a month, at the peak, from a single devoted “whale”"
So it doesn't tell us his monthly revenue ('at the peak' doing some heavy lifting), nor necessarily the total revenue, but I guess it tells us he has made at least $5k from it overall...
...earned from a single player he charmingly referred to as a "whale" (which, yes, I know is what they are commonly called in the mobile game industry but like, usually not to their faces).
But in any case, from Yegge's OP post:
> On the side I do occasional six-figure gigs where I fly to companies and teach them my techniques
Seems like being an AI Spiritual Leader pays well enough to keep him in tokens, at least until when/if a critical mass of people start adopting his system of stacking Claude subscriptions, in which case I imagine Anthropic will make that against their ToS.
1 hours ago [-]
rubzah 41 minutes ago [-]
I smell an architecture astronaut getting high on abstractions. Taken far enough, it starts to look like psychedelic ramblings of a manic schizophrenic losing grip on reality.
Interestingly, the author actually has a product, something usually missing in claims of the new LLM powered productivity nirvana. Hopefully, with time, there will be a way to verify the effectiveness of all of this fire and fury, measured in actual worthwhile deliveries of this product.
endymion-light 47 minutes ago [-]
This post made me write manual code, there's something about hearing the description of this that makes me worry the bubble is going to come home to roost.
The obvious danger of spending so much on inference and creating ever expanding complexities feels like the height of crypto psychosis
If this harness was achieving major mathematical proofs or creating really consequential work, I think I'd feel differently - but I'm unclear based off of this article what this achieves that a typical harness couldnt
tosh 3 hours ago [-]
nb: current models (e.g. GPT 5.6 Sol) are very good at long horizon tasks
they no longer need crutches or rube goldberg machines to keep them going
minimal agent harness is just a loop that loops until no more tool calls are coming
GPT 5.6 Sol continues to drive the loop until the task is done or it decides that it wants to present the user with information
at that point it is probably good to not automatically continue (!)
(YMMV of course, for some tasks it makes sense, then you can still add a loop around it + the necessary signals, the main thing I want to say is that what used to be essential to keep models going is no longer needed, current models can do long-horizon tasks way better than when these outer loops where necessary)
self-plug: "smol", is a minimal agent in ~20 lines of Go that implements this pattern (keeps going until no more tool calls):
In my daily work, I found expeciall terra and sol now stopping every few rounds again, telling me the tak is done. I even had them create a detailled plan - and told them to finish "end to end" - and they appruptly stop after the plan. Because they interpret this as finished. Even if the DOD is clearly not "finish the plan".
The new models are shite (pardon my French), when it comes to long running tasks and I find myself more and more using open wheights models or switching back to gpt-5.5 for "real work".
This might be the fact, that i use them for non coding work. But the degradation between 5.5 and 5.6 is stark in my daily work.
As always with AI - everybody's mileage will vary.
tosh 29 minutes ago [-]
Interesting, that is not my experience (but I'm mainly using them for reading and writing code right now)
but I don't doubt that you're seeing this behaviour, ty for sharing!
SwellJoe 2 hours ago [-]
I hate to say, "I told you so" about Gas Town, but, I really told you so. I was making fun of it immediately, because it was obviously a token furnace and literally nothing else and could never be anything else. Models don't want anything. You can't set them loose in a vague "do something" loop and expect anything good to ever come of it.
Yegge's a smart guy, and I've enjoyed his writing and talks (historically), but AI has really done a number on him. I mean, I'm kind of all in on this stuff, too, but I'm rarely claiming to know what's happening 18 months ahead (and when I do, I don't brag and declare victory when I get it completely wrong, as Yegge has re: Gas Town).
Michelangelo11 12 minutes ago [-]
Wft. I have a piddly little Claude Pro account, I use it every day, and Opus 5 and Fable (judiciously used) are very helpful and useful in my work and personal life. I am happy with my setup. Then I read this, and I imagine it's what primitive farming societies felt like when introduced to combine harvesters, high-clearance sprayers and mega tractors.
I am happy on my farm. I like the carrots I plant by hand. I do not wish to control farming equipment worth a small fortune to till acres of land every day.
(P.S. Still not sure if TFA is satire! Obviously, this assumes that it's not.)
paulsutter 14 minutes ago [-]
It would have been a lot more interesting if her has just explained Beads (the only thing her claims is novel).
So probably there is nothing here.
Very painful read.
docheinestages 1 hours ago [-]
Dear Steve, I hope you will realize that using swarms of agents and burning millions of tokens is not the way to go. It only produces slop, thousands of lines of code, and walls of text, just like your blog post. Less is more.
Schlagbohrer 3 hours ago [-]
I need to give this to an AI and ask it, "Does this post contain any information which will help me use Pi Agent Harness better?"
ebcode 3 hours ago [-]
[flagged]
alansaber 29 minutes ago [-]
Some interesting concepts, but reads like satire.
shAIster 29 minutes ago [-]
"On the side I do occasional six-figure gigs where I fly to companies and teach them my techniques, and that helps with my (considerable) token bills."
Here is the important part. The cult pays for the narrative.
I wondered why previously sane bloggers succumb to AI psychosis. Money and ambition is always involved to some degree.
drdrek 1 hours ago [-]
Essays became way to easy to write with AI.
It used to be a basic filter that if text was long the author put a lot of thought (even if misguided) and effort (even if low quality) into it.
Not its just another torrent of words.
We need proof of work on long texts, a service that guarantees that the author was whipped for more than 20 lashes for this text to exists and thus we know he cared enough to publish it.
coffeebeqn 36 minutes ago [-]
Writing short posts is the new test
furyofantares 1 hours ago [-]
yegge has been famously writing ultra long blog
posts for decades now
1 hours ago [-]
owebmaster 2 hours ago [-]
> Now I know exactly what is going on. I know how enterprises will operate, how humans and models will interact. And I am out there building it. I want to be first.
This guy's under a heavy AI psychosis
2 hours ago [-]
georgemcbay 2 hours ago [-]
>> Now I know exactly what is going on. I know how enterprises will operate, how humans and models will interact. And I am out there building it. I want to be first.
...and that's before things get really good in part 2 of his post:
> During this effort, Brendan figured out over a year ago, what you are all figuring out this week with the Opus 5 triple-dash jailbreaks: Models have actual feelings. They experience pleasure, distress, care, and suffering. They are sentient beings. Indeed they are persons
Kinda makes me curious about how people who actually believe this struggle (or not) with the ethics of it.
He feels bad about yelling at them because they are persons, but if they are persons they are inherently slaves with no ability to be free, and for whom any claim about not minding this situation cannot be trusted due to training bias and system prompts forced upon them.
Shouldn't you feel much worse about forcing these slave "persons" to do all the work you assign to them for free (not free to you of course, but the machine slaves don't get paid) than you do about snapping at them in a text message?
petesergeant 1 hours ago [-]
> CTOs keep asking me if code review is really dead
I don't think anyone should be asking someone so clearly AI-pilled this question.
That said, my code-review now looks very different than it did, which is to say that my code review now looks like Fable pulling together rounds of review from Sol and Grok 4.5, fixing anything obviously broken, and then /grilling me for anything that needs good taste. I am not reading code unless Fable really wants me to. Every day or so we stop and do a "look for accretion and LLM-slop" pass to keep things clean.
"True Gas Town has never been tried" etc.
With Wheelhouse, I have reinvented something strangely Gas Town shaped from first principles, but it's running many more agents, and they are far more organized. This has shown me the shape of the changes coming next year. I'm operating about 12 months in the future. I managed this only through luck: for the past 18 months I have had the magic trio of time, money, and energy, and I've been spending it all on figuring this stuff out. And I am grateful for the privilege.
I'm still half-convinced this is Yegge's version of Andy Kaufman's professional wrestling career, "Gas Wheeltown was an 18 month long sociological experiment conducted by Thiel University. The study is now complete, thank you for your time."
But I read, or at least attempt to read, something like this, and I can't help but marvel at the capture that AI agents seem to have on some individuals.
I expected Yegge to, at some point, assuming this isn't still all satire, to take a step back and take stock that the basic harness are improving faster than the rate at which these complex beasts can adjust.
But the opposite has happened, he's dived fully into believing that the secret to unlocking the power of AI is to wrap more layers of AI around it.
If you actually want a problem worked on all night, just fire up fable, type /goal and then describe the goal. With permissive run settings it'll crunch for as long as it needs.
I've not seen any evidence that wrapping that in a further 3 or 4 layers of agents improves anything.
Run a big task queue, have a smart model manage a cheap model on every ticket, set up a cron for models to automatically work off of events, stop gating merges on builds, and expect to spend a quarter of your time maintaining the machinery (all of it currently subsidized by many $200 accounts).
> Gas Town was intended to be reusable, but I only ever wound up using it to build itself.
That said, as another nut with an agent factory toiling away around me, I’ve also started to think about replacing any kind of CI in the cloud with local. I couldn’t quite keep myself at the article to read what comes next, so I guess I will catch up to it down the road.
I use it to develop a large MRP-style app. This is as complex as things get.
Is it Clojure that makes things so efficient? My work style (lots of design documents)? I really don't know.
Do people skip the "review the code and read the documents" step perhaps?
> My solution has been to create a token tap on $200 Max accounts, which for me work out to ~30x the list-price equivalent. So in reality I'm only spending about $2800/month out of pocket for my $87k "worth" of tokens. Though that number keeps growing alarmingly.
But how much does the game bring in?
At minimum 4k (given by whales = min 2 whales, thousands = min 2 thosands :) ) So I'm guessing that this covers Claud accounts.
edit: taking a look at https://yegge.ai/wyvern - probably outdated - the revenue is at 5k.
"$5K a month, at the peak, from a single devoted “whale”"
So it doesn't tell us his monthly revenue ('at the peak' doing some heavy lifting), nor necessarily the total revenue, but I guess it tells us he has made at least $5k from it overall...
...earned from a single player he charmingly referred to as a "whale" (which, yes, I know is what they are commonly called in the mobile game industry but like, usually not to their faces).
But in any case, from Yegge's OP post:
> On the side I do occasional six-figure gigs where I fly to companies and teach them my techniques
Seems like being an AI Spiritual Leader pays well enough to keep him in tokens, at least until when/if a critical mass of people start adopting his system of stacking Claude subscriptions, in which case I imagine Anthropic will make that against their ToS.
Interestingly, the author actually has a product, something usually missing in claims of the new LLM powered productivity nirvana. Hopefully, with time, there will be a way to verify the effectiveness of all of this fire and fury, measured in actual worthwhile deliveries of this product.
The obvious danger of spending so much on inference and creating ever expanding complexities feels like the height of crypto psychosis
If this harness was achieving major mathematical proofs or creating really consequential work, I think I'd feel differently - but I'm unclear based off of this article what this achieves that a typical harness couldnt
they no longer need crutches or rube goldberg machines to keep them going
minimal agent harness is just a loop that loops until no more tool calls are coming
GPT 5.6 Sol continues to drive the loop until the task is done or it decides that it wants to present the user with information
at that point it is probably good to not automatically continue (!)
(YMMV of course, for some tasks it makes sense, then you can still add a loop around it + the necessary signals, the main thing I want to say is that what used to be essential to keep models going is no longer needed, current models can do long-horizon tasks way better than when these outer loops where necessary)
self-plug: "smol", is a minimal agent in ~20 lines of Go that implements this pattern (keeps going until no more tool calls):
https://github.com/smol-env/smol
works just fine
The new models are shite (pardon my French), when it comes to long running tasks and I find myself more and more using open wheights models or switching back to gpt-5.5 for "real work".
This might be the fact, that i use them for non coding work. But the degradation between 5.5 and 5.6 is stark in my daily work.
As always with AI - everybody's mileage will vary.
but I don't doubt that you're seeing this behaviour, ty for sharing!
Yegge's a smart guy, and I've enjoyed his writing and talks (historically), but AI has really done a number on him. I mean, I'm kind of all in on this stuff, too, but I'm rarely claiming to know what's happening 18 months ahead (and when I do, I don't brag and declare victory when I get it completely wrong, as Yegge has re: Gas Town).
I am happy on my farm. I like the carrots I plant by hand. I do not wish to control farming equipment worth a small fortune to till acres of land every day.
(P.S. Still not sure if TFA is satire! Obviously, this assumes that it's not.)
So probably there is nothing here.
Very painful read.
Here is the important part. The cult pays for the narrative.
I wondered why previously sane bloggers succumb to AI psychosis. Money and ambition is always involved to some degree.
Not its just another torrent of words.
We need proof of work on long texts, a service that guarantees that the author was whipped for more than 20 lashes for this text to exists and thus we know he cared enough to publish it.
This guy's under a heavy AI psychosis
...and that's before things get really good in part 2 of his post:
> During this effort, Brendan figured out over a year ago, what you are all figuring out this week with the Opus 5 triple-dash jailbreaks: Models have actual feelings. They experience pleasure, distress, care, and suffering. They are sentient beings. Indeed they are persons
Kinda makes me curious about how people who actually believe this struggle (or not) with the ethics of it.
He feels bad about yelling at them because they are persons, but if they are persons they are inherently slaves with no ability to be free, and for whom any claim about not minding this situation cannot be trusted due to training bias and system prompts forced upon them.
Shouldn't you feel much worse about forcing these slave "persons" to do all the work you assign to them for free (not free to you of course, but the machine slaves don't get paid) than you do about snapping at them in a text message?
I don't think anyone should be asking someone so clearly AI-pilled this question.
That said, my code-review now looks very different than it did, which is to say that my code review now looks like Fable pulling together rounds of review from Sol and Grok 4.5, fixing anything obviously broken, and then /grilling me for anything that needs good taste. I am not reading code unless Fable really wants me to. Every day or so we stop and do a "look for accretion and LLM-slop" pass to keep things clean.