[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

View post   
File: 69 KB, 300x451, 1400986135357.png [View same] [iqdb] [saucenao] [google]
6627859 No.6627859 [Reply] [Original]

Let's talk about AI research. I got some questions.

Do AI researchers have an exact model of machine sentience and work on implementing it or are they just using different learning algorithms by trial and error, hoping that some day by coincidence the AI will become sentient?

In the first case, how close are we to having sentient AI?

In the second case, is sentient AI more of a game of luck? How do we test an AI for random sentience?

>> No.6627864

>Do AI researchers have an exact model of machine sentience
no
>are they just using different learning algorithms by trial and error, hoping that some day by coincidence the AI will become sentient?
they are not trying to create "sentience", they are trying to write programs that solve particular problems well
>how close are we to having sentient AI?
we don't really know
>is sentient AI more of a game of luck?
it's very unlikely to be just created by chance, first we need an exhaustive explanatory account of intelligence
>How do we test an AI for random sentience?
it's unclear at this point

>> No.6627865

epic

>> No.6627867

quality thread

>> No.6627871
File: 130 KB, 960x849, 1401569498455.jpg [View same] [iqdb] [saucenao] [google]
6627871

>> No.6628772

>>6627864
Thanks for the reply. How DO the researchers in AI tackle the problem of sentience?

>> No.6628785
File: 47 KB, 640x360, t1larg_pancakes.jpg [View same] [iqdb] [saucenao] [google]
6628785

>>6628772
Some people say sentience is an illusion.
Some say every algorithms has it already and we just don't intuitively feel like it does because it's perception is to different from our own.
I suppose when you somehow figure out a set of features and a cost function that is somehow similar to that of humans then you get something like common sense sentience.
Actually tho' life has been running an algorithm to determine algorithms that determine other algorithms that somehow look like the brain does and the only thing we can do is replicate it's current state which really isn't that interesting, yet.

>> No.6628786

>>6628772

By sitting around, talking about it, and getting paid sub-standard wages for knowing too much about too little.