I first made a prototype of the core mechanic back in 2020. For a while I did not have too much time to dedicate to the project, but the idea had been in the back of my mind ever since.
After befriending Tobi, who now contributes to the game’s development, and performing several playtests with very positive results, I’ve decided to commit to the development of Botnet of Ares.
Botnet of Ares is meant as a critique of security practices in some consumer electronics and IoT devices. It is also a response to the excitement & anxiety many of us feel regarding recent developments in generative AI & large language models. Real life botnets are very much a thing, and so is automatic exploitation of such devices. Botnet of Ares asks the question: what would happen if an AI Agent were trained to exploit and uncover vulnerabilities across the entire Internet?
I want to be specific with my language, as the AGI Singularity replace-all-knowledge-work AI hype is still pretty far from what we have today: large language & multimodal models and other types generative AI.
For me it’s been more of a time-saver than anything else. For example, large language models are quite good for certain repetitive refactors - faster than coming up with a Vim macro! I’ve also found large language models useful for working with… Language! They’ve been great for coming up with evocative names for things, or rephrasing written text to have a different tone of voice. They’re also good at extracting specific information from huge documentation (e.g. using RAG) but of course you have to check the details yourself as these systems can be wrong. Rather, it saves the time you would’ve taken to find what you’re looking for on your own.
I haven’t had much luck using them for big programming job, but I can definitely see it saving time. And I don’t think the models we have today can create production-ready creative writing, but they can save a writer quite a bit of time!
I wouldn’t really use generative AI for graphical assets - it’s hard to get a consistent look, and a lot of people think that it makes certain products look ‘cheap’.