It’s not an agent with its own goals so in the gamedev definition, no. By calculator standards, also not. But just as a washing machine with sufficient smarts is called intelligent, so it’s, in principle, possible to call a calculator intelligent if it’s smart enough. WolframAlpha certainly qualifies. And not just the newfangled LLM-enabled stuff I used Mathematica back in the early 00s and it blew me the fuck away. That thing is certainly better at finding closed forms than me.
It’s literally peaks and valleys of probability based on linguistic rules. That’s it. It is what’s referred to as a “Chinese room” in thought experiments.
This is not claiming machines cannot be conscious ever. This is claiming machines aren’t conscious right now.
LLMs are like databases with a huge list of distances allowing you to find the “shortest” (aka most likely) distance to the next word. It’s literally little more than that.
One day true AI might exist. One day perhaps… But not today.
Removed by mod
I don’t doubt the possibility but current AI tech, no.
It’s barely even AI. The amount of faith people have in these glorified search engines and image generators Lmao
I don’t have a leg to stand on calling anything “barely AI” given what us gamedevs call AI. Like a 1d affine transformation playing pong.
It’s beating your ass, there, isn’t that intelligent enough for you?
A calculator can multiply 2887618 * 99289192 faster than you ever could. Does that make a calculator intelligent?
It’s not an agent with its own goals so in the gamedev definition, no. By calculator standards, also not. But just as a washing machine with sufficient smarts is called intelligent, so it’s, in principle, possible to call a calculator intelligent if it’s smart enough. WolframAlpha certainly qualifies. And not just the newfangled LLM-enabled stuff I used Mathematica back in the early 00s and it blew me the fuck away. That thing is certainly better at finding closed forms than me.
It’s literally peaks and valleys of probability based on linguistic rules. That’s it. It is what’s referred to as a “Chinese room” in thought experiments.
This is not claiming machines cannot be conscious ever. This is claiming machines aren’t conscious right now.
LLMs are like databases with a huge list of distances allowing you to find the “shortest” (aka most likely) distance to the next word. It’s literally little more than that.
One day true AI might exist. One day perhaps… But not today.