Join the discussion
Write your take first — we'll ask for email only when you're ready to publish.
- Hacker News
- Holy moly https://yegge.ai/essays/model-welfare/ ! Talk about spicy takes. Fable is sentient, be nice to it, or in six months I won't be your friend.
fwiw, i give it a 1/3rd chance he's right.
- I give it a 0 chance he’s right.by shric
- 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?
by jwr - I'm on similar plan through work ( Teams max ), and I've only ever reached the limits by backgrounding large exploratory feasibility studies while also getting on with actual work.
From what I can tell however, true "vibe coding" gets a lot more expensive very quickly.
If you're actually just agentic engineering, then after a couple of mis-fires you'll stop, re-analyze what it is doing, what you asked for, and consider different approaches.
With pure "vibe coding", you'll keep asking it to meet your requirements as it twists and turns in the background to try to fit.
A more concrete example from a personal "game" I'm developing as a hobby:
Fable got itself in knots trying to render some fairly basic blocks on top of a wall. It understood the problem, it consistently rendered as "fins" under screenshot, and thrashed away for a good while trying to produce better looking blocks to no avail.
Eventually it stopped, and lied that it was happy with how the walls now rendered.
It took me 30 seconds in-game to realise that it was only rendering backfaces, once clued into that, the machine was able to trivially fix it without using much tokens.
But then another similar issue popped up later, hedges exhibited the same behaviour.
It would have been easy to accept the solution to render the hedges in the other order too, but I demanded a better explanation, and it turns out that it had fairly early on defined a DrawBox() method that was backward to Unity's expectations, and pushing further, it emerged this all started when I asked it to transfer what it had learned from a webGL prototype to unity.
It took things a little too literally, which is also how I ended up with an Eastward sunset at first. When I asked it to fix that, I hadn't realised that it's solution was to essentially reverse coordinates and world understanding.
Now, this stemmed from a lack of understanding or review of the 3D code, it's a vibey hobby that I'm using to learn what really goes into rendering a scene.
Had I been a professional, I would have absolutely caught that error ( and many others ).
So being professional would have helped reduce token expenditure massively, firstly by slowing down what could be done to leave time for review, but secondly and perhaps more importantly, by avoiding it getting into a state where it thrashed away for an hour trying to solve an impossible puzzle because the base assumptions were wrong.
by xnorswap - Automated AI factories destroy tokens. 2 codex and 2 Claude subs on mine. A lot is waste and I’m looking for optimizations. Like using pi instead of CLIs from the frontier labs is 1/10th the context.
- The article talks very directly about abandoning human code reviews. Couple that with a layer of AI agents whose purpose is to produce work for other AI agents, and you've got yourself a very efficient token burning feedback loop.by rco8786
- The guy built an engine designed to burn tokens: https://miro.medium.com/v2/resize:fit:1100/format:webp/1*85x...
Each little critter in there is a component that both consumes and produces tokens, and it runs in a feedback loop, so each token inserted into the machine will cause it burn exponentially more tokens in order to output yet another token.
Ostensibly its purpose is to create software, but what it does is very efficiently burn tokens as fast as possible.
by ModernMech - 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):
https://github.com/smol-env/smol
works just fine
by tosh - Forgive me if a dumb question, but is there no compaction? Do you ever run out of context?by movpasd
- 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.
by sdoering - 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
- This is what he's spending '$87k worth of tokens' building a month btw: https://store.steampowered.com/app/1541710/Wyvern/by harmoni-pet
- The reviews (both the contents and total count) are not exactly confidence inducing.by rco8786
- I watched Steve talking about GasTown in some podcast but I didn't understand it at all. I read some part of this article didnt understand what he wanted to say.
I still remember his very old blog post where he rants about Google and Bezos, after reading that I thought the author of this post is very insightful but when I read his post I have no idea what is he talking about.
by sumedh - His game has only a handful (like literally 5-8) players [0], at least on Steam.by redhale
- and he hallucinates that he is making a lot of money from it, direct payment from the users. I played the game, it is so bad.by brazukadev
- The game was first released in 2001, but released on steam on 2021. Needless to say, most players don't play through steam.by sapiogram
- 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.
by xnorswap - I call it "yegge sweats"by nautilus12
- MOE and COT or github co-pilot. I think agents can marginally improve the performance of AI.by deadlast2
- 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.
by xnorswap - 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).
by ramoz - 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.by cmpxchg8b
- > 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?
by pringk02 - Keep in mind median lifetime game revenue on steam is somewhere between $250 and $1000 depending on how you slice it...by jezzamon
- I tried playing it. It's... not going to change the world of gaming.by EdwardDiego
- > 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.
edit: taking a look at https://yegge.ai/wyvern - probably outdated - the revenue is at 5k.
by mad0 - 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.
by rubzah - 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 ;-(by bambax
- > They were all way too long
And still are. When I got to his "Lots to cover today! Let's get to it.", I was like "wait, we haven't started yet?"
I used to read his stuff regularly prior to his Sourcegraph days, and I don't recall why I stopped. Completely agree that he was always fun to read.
So I was really disappointed to start reading this one (I'll leave the tab open, but probably won't finish it). Sounds like he's pretty out of touch with reality these days and isn't grounded at all.
Like the sibling poster, I gagged a little at his casual mention of six-figure teaching engagements, and the "suckers" (that's what "whales" read to me as well) keeping his game alive.
by kelnos