Reflecting on Bitmagic's year as the Lightspeed Game Changer in AI Agents

If last year was the year when LLMs became mainstream, this year has been all about AI agents. It's truly amazing how much can change in a year. When Bitmagic won the Lightspeed and GamesBeat Game Changers 2025, we were the standout winner in Generative AI and Agents category, but I remember that "and Agents" was added to the category name in the last moment!

What we originally were nominated for was "Generative AI" because AI Agents were not a thing yet when the contest started earlier in the fall.

A year ago, Moritz Baier-Lentz 🎮 and Dean Takahashi obviously could see the future. Back then, Bitmagic wasn't using agentic AI coding. I remember being slightly amused for being the game changer in AI Agents category and not even embracing the tech. But little did I know.

Now, just 12 months later, Bitmagic V3 is truly a game changer in how games are made. It's currenly out in an invite only mode, so only a very few people have ever seen it, but those who've used V3 know what I mean.

It's a 100% agentic "vibe coding" platform that helps people create innovative games that are great to play. As I'm typing this, the text editor on my Mac warns that agentic is not a word, perhaps I meant "agent" or "agents"?

One year ago, Bitmagic was running on Unity. Now we have our own game engine, written from scratch, designed for AI. Again, a game changer. This is how things will be one in the future as it enables so many things that were not possible before. Plus, it helps us iterate super fast. When a researcher puts out a paper about a new innovation and we feel it's important, we can have support in Bitmagic for it the next day.

An interesting side effect of the accelerated pace of innovation is that since the speed of fundraising process has largely not accelerated, the opportunity cost of startup fundraising is becoming a bottleneck.

Fundraising has traditionally been a full time job for the CEO. Now that team are smaller and the CEO can be an important part of the development team, putting everything aside to focus on the next funding round is too expensive.

This year I've greatly cut down the time I spend with prospective investors. I'm cutting it down to focus on a small group of investors. Gone are the days a CEO could go all out and pitch fo 100 VCs. You can pretty much build a product during the time it takes to do that. I don't know if this is controversial, or is everyone doing the same?

I'm noticing that I spend some of that time to be more visible on LinkedIn. I post much more technical articles of what we're building than I used to do. Part of the reason is to attract talent, but an important side effect has been that I'm starting to have deeper discussions with investors outside of fundraising cycle.

This strategy seems to be paying off and the number of the quality of discussions I'm having both experts and investors in the AI industry is up dramatically. I'm curious to see where this leads but so far Iike this new future.