I wrote a tested prototype MMO in the last few weekends with AI as my power tools.

You're holding it wrong.

I'm VERY interested in your project. Not playing it, I mean more the techniques and tech stack. That sounds entirely out of my reach with an AI, and ive written game engines in c++ before. Networking, synchronisation problems, etc are really really hard.

What process did you use with the AIs, any prompting insights - context, agentic prompts etc?

What tech stack did you use that you found AIs were familliar enough with, ive found them woefully misinformed about most libraries and technologies ive tried them with in game development, often confidently mixing out of date and new information

I'm using Cursor's composer agent mode with Sonnet 3.5 (I don't use OpenAI on principle, snakes). It does a great job of finding the relevant code without overloading its context window.

I experimented today with Aider (to get R1 involved) and had less success, but it might be that I don't have the workflow down.

I have found cursor can handle a .NET C# back-end using highly standard code structures very well. SignalR for networking.

I've created servers and very basic HTML visualization for three projects - a fairly simple autobattler (took a day), a web-based beat-em-up (2 days), and now a bit more ambitiously my dream RTS-MMO (3rd weekend running).

I started with concise MVP specifications including requirements for future scaling, and from these worked with the AI to make dot point architectural documents. Once we had those down I moved step-by-step, developing elements and tests simultaneously then having the agent automatically run the tests and debug. The test-driven debugging is the part that saved the most frustration, as the initial implementation was almost always broken, but leaving the agent to its own devices (tabbing in and typing "continue" when hitting Cursor's 25 tool call limit, sometimes for hours) the tests guided bug fixing and amazingly it got there fairly consistently, though occasionally it will go off the rails and start modifying the tests to pass or inventing unwanted functionality.

The code is as standard as possible, with the servers all organized identically API -> Application -> Domain <- Infrastructure, and well separated between client/server. Getting basic HTML representations wasn't an issue, but it does begin to struggle and requires a lot more direction when it comes to client-side code that expands beyond initial visualization. I had a lot more success with Monogame C# than Phaser or other web formats (e.g. I quickly gave up on SFML, same issues you were having).

I'm a professional game developer but without formal CS/programming training, so I'm aware of my requirements but not always how to implement them cleanly. I understand the code it writes which feels vital when it occasionally rolls a critical miss, but these projects would have taken me months without AI.

How so? There are so many tangible applications already: reduced customers service costs, legal research, analysis of medical records or imaging, self-driving Waymos, and so on. Just the things I listed will have profound impacts on cost savings, productivity, and quality of life.

Little Boy certainly had "profound impacts on cost savings, productivity, and quality of life" in Hiroshima.

Impact isn't inherently positive.

How so?

When everyday people are telling you their opinions about what AI is gonna do you know it’s a fad.

Everyday people were talking about iPhone. That was what made iPhone a fad and prevented Apple from being the company with the highest MarketCap.

iPhone was a fad. Everyone and their grandmother had an idea for some kind of app. It was all people talked about, apps apps apps.

Then the market blew up and a few big winners ate up all the profit and everyone else died on the long tail.

Today, iPhone exists, but people don’t even think about them much. They just use it.

Pestering someone now about your iPhone app idea is like pestering someone about your website idea back in 2015.

Everyone and their dog keep trying to force me to use their apps, like there's a mantra inside their marketing divisions that apps are the shit. Doesn't look like apps are over. The market is just saturated of "app for X". The hype of getting rich from zero through one app is over, but apps aren't.

VR movies was a fad. You can date a TV by it's "VR mode" feature on the remote to a few years. No one is trying to sell me VR TVs anymore. That's what a fad looks like.

What you’re talking about really isn’t that iPhone is or was a fad. What’s really happened is innovation has essentially halted and it has become a commodity like all smartphones.

Every day people were doing that with the Internet.

Yea, and then came the dotcom bust.

Yeah, and the Internet just never recovered!

I wish this whole internet fad would just die already!

I'd write a script that crawls all these AI topics and use an LLM to count how many times this conversation has been had but I'm too tired to do so lol. Just don't feed the troll / dumb person. If they're too stupid to see how revolutionary LLMs are by this point.. just let them be stupid. Just downvote and move on with your life

We’ll know when it can park a car in an everyday parking spot without messing up your Grandma’s Camry I suppose.

Oh yeah, let's all wait till then to get the value out of models today.

I’m not saying it doesn’t have value, but it’s not worth my time to spend 20 minutes to prompt engineer a tool to write me a plagiarized document I could write myself in 20 minutes? Why would I invest my time into using a tool that undermines my own value? What’s the value prop for me?

Your inability to find value propositions and use cases is your issue.

K, just keep this conversation in mind in a decade or two when you realize that your input is the product, not what you got out of it.

Oh yeah—tell that to my AI pipelines on my local compute.

Local pipelines are great and all for now but there’s practically no way those will be able to keep up with server based models long-term. If those are useful for you today, that’s great.

Right and those "server based models" can be deployed by me or my team, on clusters we own.

You're talking in a thread about Deepseek...

It certainly lets untrained people create stuff they probably should not

Only if you’re moving the goal posts every other week.