I regularly use Google Lens and I suppose it is pretty impressive at recognizing stuff.
I've never really looked into the companies that offer to do the training, maybe there's something going on there.
One annoying application of this tech is more sophisticated bots on the internet writing generated articles, messages, comments, etc and thereby making it harder to find real humans online.
AI still hasn't been able to generate valid software for any meaningful purpose. That would require the AI to vaguely understand the problem, translate it into procedural steps, then iterate over those steps rewriting parts that don't work.
So as long as it can't generate real low level code (machine code or assembly, not some highly abstracted API), the technology has reached it's limit.
In fact, the mathematics to do this stuff has been around for a while, but the only reason it's taking off now is because processors have gotten exponentially faster. (the rate of speed increases has been slowing down as well, so unless a new computer architecture is introduced or quantum computing works, that's basically the end of it.)
What do you guys think about Neural Networks? Are they really going to change the world, or have they pretty much hit their limit? At this point I think it's all about the APPLICATIONS and machines they'll be hooking the AI up to
The world is a giant google datometry of everything. Thanks for breaking IT down into humanely digestible pieces. The only suggestion I have about this piece is to replace the words conspiracy theory with objective analysis. Who buzzed the phrase “conspiracy theory”? I can smell nefarious intentions every time I hear the term. I think it is injected to propagate “settled” science, to reduce possibility and analysis in to a refutable term that can be swatted. By refusing to use the term theory, we could start to retrain the nearal network modelling system.
I regularly use Google Lens and I suppose it is pretty impressive at recognizing stuff.
I've never really looked into the companies that offer to do the training, maybe there's something going on there.
One annoying application of this tech is more sophisticated bots on the internet writing generated articles, messages, comments, etc and thereby making it harder to find real humans online.
AI still hasn't been able to generate valid software for any meaningful purpose. That would require the AI to vaguely understand the problem, translate it into procedural steps, then iterate over those steps rewriting parts that don't work.
So as long as it can't generate real low level code (machine code or assembly, not some highly abstracted API), the technology has reached it's limit.
In fact, the mathematics to do this stuff has been around for a while, but the only reason it's taking off now is because processors have gotten exponentially faster. (the rate of speed increases has been slowing down as well, so unless a new computer architecture is introduced or quantum computing works, that's basically the end of it.)
What do you guys think about Neural Networks? Are they really going to change the world, or have they pretty much hit their limit? At this point I think it's all about the APPLICATIONS and machines they'll be hooking the AI up to
The world is a giant google datometry of everything. Thanks for breaking IT down into humanely digestible pieces. The only suggestion I have about this piece is to replace the words conspiracy theory with objective analysis. Who buzzed the phrase “conspiracy theory”? I can smell nefarious intentions every time I hear the term. I think it is injected to propagate “settled” science, to reduce possibility and analysis in to a refutable term that can be swatted. By refusing to use the term theory, we could start to retrain the nearal network modelling system.