It's always good to see that the newest technologies get involved in the use cases of new projects. What I personally believe is all the projects should have saved space for upcoming technologies so that they can keep upgrading for good. However, it doesn't change the fact that not all projects bear fruits, there could always be shit projects. Calculated risks, thorough fundamental analysis, etc. can help manage the risks involved with investing in any project.
You have just said it right.
Many are intriguing and interesting projects whether they are AI related niche projects or not. What matters most is how you're choosing a project being an investor.
These projects will always be there and you'll just need to make your own decisions rightly and thoroughly. Apart from that, it is your money that should be invested well.