Sentience and intelligence are vastly different things.
As it stands, these systems have yet to make a compelling case for intelligence.
As it stands, these systems have yet to make a compelling case for intelligence.
Given that we don't even fully understand how the human brain works we can't even specify how close or how different these things are to the way we think.
Articles like this aren't really open minded. How can you claim AI is a mirror when you don't even fully understand the thing that it is reflecting?
You can't know if you're a mirror yourself.
Why would one assume that intelligence existing on a different substrate would be artificial?
So far we have no example of intelligence that is not tied to a living organism.
If a computer can simulate a living organism to a sufficient degree of detail the simulation may be intelligent and alive.
Thus life is at least a sufficient condition for intelligence. It may be a necessary one as well.
How GPT is different from life is that it lacks biological goals. It also lacks the abstract modeling ability that higher organisms have.
In other words GPT fails because it relies on combinatorics without real modeling of the world (aka understanding).
Unless you can give such a definition of intelligence, then you are in no position to assess whether these systems are intelligent (or in what way, to what extent) or not, even to the extent of meeting your own criteria whatever that may be!
There's a lot of people over hyping this, but there is an even larger group of people downplaying everything and refusing to believe that a paradigm shift in AI has just occured.