Upgrade your account to a Community Supporter account and remove most of the site ads.

Reply to thread

Message: <blockquote data-quote="Stalker0" data-source="post: 8868466" data-attributes="member: 5889">General AI is really not necessary. We can do it using the "toolbox" approach, creating very specialized AIs to do very specific things. But you have enough of them, and suddenly they are handling everything.I could make a impressionist Art AI, or pick another art style AI (hehe not an art history guy <img src="https://cdn.jsdelivr.net/joypixels/assets/8.0/png/unicode/64/1f609.png" class="smilie smilie--emoji" loading="lazy" width="64" height="64" alt=";)" title="Wink    ;)"  data-smilie="2"data-shortname=";)" />. You could make an AI for every specific style you can think of, as well as an AI that just thinks of new style. You don't have to have one super AI.When you say gradual, I think you also seriously underestimate the pace of progress. We aren't measuring success in decades, they are in years, sometimes in months. The power of the machine is its speed, getting it to do something is very hard, but once it learns to do something well it takes off like a shot....especially with software where you don't have to pay for it to make copies of itself. The replication is relatively "free".Now that we have art AIs, you are going to see rapid rapid progress in their development, measured in months, maybe years, but not longer than that.Further, we have to understand the notion of scope here. Once you have an AI good at a certain kind of decision making for example, that doesn't affect just one industry, it effects dozens. Change will not be that gradual once this ball really starts rolling down the hill.</blockquote>

Verification