Attention Is The True Bottleneck

And I don't mean the so-called attention mechanism inside the architecture of a large language model.

What I mean is the very limited attention that a human audience can devote to apps, images, songs, or movies. Video games are a great example here because they combine hard software engineering with a lot of artistic expression. AI can speed up both. Faster coding, easier time generating the various visual assets. Will we therefore see an abundance of successful video game projects?

I don't think so. Things might shuffle around a bit, in that smaller studios might be able to punch above their weight. But from the consumer's point of view, it's not like we have a shortage of video games to choose from: There are over 100,000 games available for purchase on the Steam gaming platform. And if you can prompt an AI to "make me a game and make it good", so can everybody else. The current situation, where a few titles get most of the sales and the rest languishes, will remain in place.

Hard things are hard and there are no shortcuts. If a new tool makes aspects of the hard work easier, it will lift everyone and expand the boundary of what's now hard but no longer impossible.

The positive in this: If you're taking your craft seriously and are constantly pushing against the boundaries of what's possible, AI will be less of a threat. The difference is between

"AI makes this easy, so I can just phone it in"

and

"AI makes this easy, so now I can put even more energy into that other part"

And those who do the latter will get all the attention.