I built https://gallotails.com/ because I am a cocktail enthusiast. First commit, March 6th 2019. One of the features I am proud is which cocktails you can build based on your ingredients, and I was smart, the bar ignores garnishes and can suggest cocktails you're one ingredient away. But you have to add the ingredients. And interpret the bar screen.
And here's what I did with ChatGPT a few days ago: https://chatgpt.com/share/6ab59877-4b74-83e8-8da5-966d9f3a38...
Clicked the microphone icon and rambled for a few seconds (could have just taken a picture too I guess). Nudged ChatGPT to only give me simple cocktails I can stir instead of using a mixer. Said I could buy a lime, sure. Then asked how to keep the ginger beer fresh.
There is NO WAY I can code all of that in my little gallotails.com - a single chat window has entirely replaced my site. And that's my pet little project I do it on weekends, mostly to keep my tech skills sharp. I can only imagine what the biggest cocktails website, actual companies, are going / will go through once more and more people just realize the chat tab is enough. And I am not talking about purely content, ChatGPT can do everything my site does, and more, into any direction, in an instant.
Which left me with the question, then what should my site be? Community? To be taken over by agents?
Currently munching on what sites like mine should do, and what they should be.
Oh yeah, and I am not mourning or anything like that. And I do want the human touch, because AI can't. But humans, at least over the internet, can send their agents, or companies can run agents and I won't know from my site if it's an actual human, or some AI fabricated interaction (can't even trust an image it uploads).
When AI first showed up a few years ago, I thought about opening an actual physical cocktail bar. Hypothesizing humans will be tired after a long day of their agents talking to other agents, they would crave for actual humans. I would ban electronics inside - no distractions. Maybe one day I will go for that...
This is not sustainable in the long term, but irrational longer than solvent yadda yadda.
Chatgpt would shit on my great grandma's waffle recipe but it's the one I'll keep making till the day I die.
Eg from your share, the model output “I prefer it Scotch-heavy rather than 50/50 because otherwise it gets very sweet.”
It’s basically cribbed this from someone; it literally can’t taste.
The frontier labs are not yet able to do rlvr over mixology/meatspace. So maybe think on how to capitalize on that.
My software development process feels exactly like every debugging session in the holodeck that seemed terribly unrealistic to me.
I'm in SRE so systems reasoning is something of the value. But me writing code or even configuring stuff? That's dead. It is a total waste of time - Claude can do it better, faster and it works.
- Video devlogs were long a shitty way to make a community (I can explain this one in more detail, but in short, the tradeoff between opportunity cost and video quality is a Pareto curve that is bad at all points on the curve, if your goals include both “make game” and “build community for players with devlogs”, the way out is to drop one of those two requirements, either “make YouTube/Twitch/TikTok channel” replaces “make game”, or you build some other audience for your devlogs)
- Good devlogs don’t have that level of detail, to let people easily recreate games
- Most games can’t easily be replicated by LLMs, in short, the people who are good at steering LLMs like that are making their own games.
This is not the first time I heard this story. Sometimes it comes with the lamentation “back in the day people could build communities with devlogs” and no, that was generally not a good way to build communities. It was mostly good streams from YouTubers who were making the game in order to make YouTube content or it was mediocre streams from random devs. Throw in a few people who are famous game devs who choose to stream and get a big audience because they already have a community.
The whole problem with LLMs is that you don't need the detail.
An LLM "recreation" doesn't have to be complete or very good. It just has to steal enough thunder to be profitable.
It's it possible for an LLM-created game to be profitable but "unsuccessful?"
For instance, if (on average) if it costs $100 in tokens and time to make a crappy LLM-game clone, but you can clear $200 in sales per game on average, you're ahead.
And the economics of slop mean there are a lot of people in 3rd world countries who will make all that effort for a $100 payout (or less).
My point is they don't actually have to do a good job, they need to make a pile of shit that looks just good enough to trick a few people into buying it.
Scenario one: they’re competing with my game and stealing my thunder. In this scenario my game wasn’t very good to begin with.
Scenario two: we’re going after different customers. They get the customers who want a cheaper game, I’m getting the customers who want a better game.
And someone trying to vibe copy another person's game is definitely the type who has no idea of value to add. So many of the slop games the whole concepts are so generic that they definitely just asked chatgpt for everything.
In the modding scenes for some games I play I've seen vibe coded mods where the gameplay additions make no sense and have no sense of balance or fun, with these completely new to the community devs having ko-fi links set up from the start.
AFAICT this is a fear that some game developers have, that someone will steal their game and run with it, and the stories told are more repeated based on this shared fear than based on realistic scenarios. People repeat it because it’s a good story.
And yes, I’m aware of some situations like 0x10c and the like.
The idea is not valuable, ideas are a dime a dozen. The actual value is in your implementation.
There's no moat in an idea either so I guess it's down to marketing budget.
It requires discernment, taste, on the part of the user to be able to point the LLM at the weird parts that look wrong.
Even with that, the LLM can only fix most of, not all of, those rough edges.
More specifically to AI... A blog on a nice website is no longer a signal for quality so there's just no point in devs investing in that.
It really makes me appreciate how important good game design is. Claude is doing a fine job coding everything I describe, but it doesn't really understand fun, so I need to.
(It's not a great game)
And that's why we can't have good things...
(This is just gonna keep happening more and more until eventually we'll need something like a patent system for ideas)
So, a patent?
> That's why I created isitsaas.io: your platform for determining whether your llm wrapper has product potential, or whether people would rather just use claude directly instead. Simply pop in your business idea, and our award winning, proprietary technology will ask claude if you can make money off of it.
> Trial memberships start at $10/month
Then it’s sufficiently non-obvious. I believe there’s still a pretty big field of things that fall into this category. Making proper things is a lot more than writing some code.
I've been playing with an LLM backed reference checker for my partner who is a lecturer. For each reference in some work she's marking an agent is dispatched to read the referenced source and check that the reference is correct / accurate / not hallucinated.
It's useful to her, but it feels like it would be a waste of effort to turn into a product, because pasting the same document into Claude/ChatGPT with a prompt like "download and read each reference etc" would work pretty much just as well.
I had similar concerns about an AI backed training generation company I interviewed with. I asked how they saw themselves competing with increasingly capable generic harnesses. I don't recall exactly what they said but the impression I got was they were so focused on competing with the big LMS players that they hadn't really considered it.
If you don't care so much about adding something significant to the world, and your LLM-buttons-wrapper is a novelty, chug along and see where it goes if you're having fun. Looking for what value mean at this moment is a much harder effort, and probably a luck-based endeavour.
I don't have fun creating LLM-wrappers or at least haven't thought of a fun idea for it that could be even a fun novelty to work on so personally I'm not doing it but I'm using LLMs for other fun stuff that required much more of my free time before.
Very few things can't be reduced to insignificance. MacOS was just cribbing PARC, Facebook is just a glorified PHP forum, Dropbox is just SFTP, etc. etc.
Even in deep-tech, Zipline is just wrapping from deeper-tech (batteries motors etc), GLP-1s were a VA throwaway that dusted off, etc. etc.
LLM wrapper doesn't mean anything, it's an implementation detail. Besides being an LLM wrapper what is a given thing?
The beer app wasn't just a beer app, it was the intersection of the first time accelerometers were doing something that the average consumer could interact with in their pocket, the first time there was something to spend money on for your phone besides a wallpaper, a ton of things.
If anything now when something seems trivial or like a toy, but it has a lot of traction, I want to spend energy figuring out why is it more meaningful than it appears.
I don't get what exactly you are talking about as a reply to my comment, what exercise exactly? Deciding if something is worth spending time if you're having fun with it?
> If anything now when something seems trivial or like a toy, but it has a lot of traction, I want to spend energy figuring out why is it more meaningful than it appears.
Sure, if you're having fun (and in fun I include a sense of accomplishment, curiosity, whatever tickles you) with that, go for it.
It's a strangely arrogant reply to my comment, perhaps you read something into it that I didn't say and replied to that?
Also I honestly thought we were two arrogant people conversing in good faith: you don't find unarrogant people saying things like "If you don't care so much about adding something significant to the world..."
I think text distorts what I meant with that, I personally don't care much about adding something significant to the world, not anymore at least.
I found my peace by pursuing what interests me personally even though I know it might not be of much significance to anyone else, and that I won't be pursuing the path of "adding something significant" just for the prestige of it in my life.
In a sense it's the opposite of arrogance since I care much more that the person is enjoying the process of whatever they are doing rather than the results of it. I've seen too many bright friends end up miserable chasing it, I've felt too miserable myself when putting too much focus on that, I'd rather have people enjoying doing what they are interested at if it's not something harmful.
Start from thinking about product value add, and user behaviors first.
Also for creatives, some sort of aesthetic control? Most of the examples on this page look like templates you’d see in an office suite that appear swish, but are ultimately bland.
You can also host a chat interface that connects to said MCP server as a convenience. But serious users probably already have their own inference and it's already hooked into other resources as well.
The LLM only does average returns from average inputs.
Some SaaS products have network effects, but that doesn't appy to new software.
Starting a SaaS business in 2026 seems a bit silly to me. Not saying people won't make money, but you'd have to be pretty lucky.
Nope. If all your doing is prompting an agent to build stuff (especially if what it is building is based on LLMs), then you might as well take it back to the whiteboard and rethink how you're spending your time/tokens.