One category that seems to be missing is speech - the quality and speed of voice cloning has come a long way in the past year. Also, I agree that 2024 is probably going to be even faster - I'm not sure when things will slow down, but we've probably got at least a couple of years of low hanging technological fruit to zoom through first.
I expected someone to point that out! I did toy with the idea of including it, but since that wasn't a category in my original article, I decided to skip it. (Although my original article did talk about Whisper's impressive speech recognition.) No doubt there were crazy advances in text-to-speech, and MyShell's newly released OpenVoice will likely be in tomorrow's 10X AI. It's scary how little voice input is now needed to faithfully clone a voice.
I actually think we'll see a funny thing where the pace of development increases but investment might actually slow down. I think the space was so overhyped and so many startups burned through cash without actually making a profit that there'll likely be a bit of a cooling. But I may be wrong. (It's been known to happen.)
It's tempting to draw comparisons to the 95-99 period, when the internet frenzy really took off. If I had to guess, I'd say that the 4 or 5 year dot-com mania will be compressed into like 2 years. There's always so much sloppy investment whenever the taps turn on like this.
One category that seems to be missing is speech - the quality and speed of voice cloning has come a long way in the past year. Also, I agree that 2024 is probably going to be even faster - I'm not sure when things will slow down, but we've probably got at least a couple of years of low hanging technological fruit to zoom through first.
I expected someone to point that out! I did toy with the idea of including it, but since that wasn't a category in my original article, I decided to skip it. (Although my original article did talk about Whisper's impressive speech recognition.) No doubt there were crazy advances in text-to-speech, and MyShell's newly released OpenVoice will likely be in tomorrow's 10X AI. It's scary how little voice input is now needed to faithfully clone a voice.
Whew, fire hose for sure.
Is it fair to say that we saw twice as much change in 2023 as we saw in 2022, which was a revolutionary year already?
Tell me about it. At this rate, the next version of this post will be a novel!
Will 2024 be twice as consequential as 2023? Maybe we can quantify some of this disruptive change with investment dollars.
I actually think we'll see a funny thing where the pace of development increases but investment might actually slow down. I think the space was so overhyped and so many startups burned through cash without actually making a profit that there'll likely be a bit of a cooling. But I may be wrong. (It's been known to happen.)
It's tempting to draw comparisons to the 95-99 period, when the internet frenzy really took off. If I had to guess, I'd say that the 4 or 5 year dot-com mania will be compressed into like 2 years. There's always so much sloppy investment whenever the taps turn on like this.