Because you're using it wrong.

Really. Tool use is a big deal for humans, and it's just as big a deal for machines.

Wouldn't an intelligent computer know to use tools? The core of the point being discussed seems to be why do you need to ask it to make it you inventory software when an intelligent system would know that when asked to build an inventory system setting up a database and logging all the information is need and ask agents to do that.

▲

CamperBob2 2 days ago | parent [-]

It's the same question that you might have asked in 1920. "This radio hardly works at all. Can't they do something about all the static? I don't see the big deal. This is just a scam to sell batteries and tubes."

▲

roommin 2 days ago | parent | next [-]

I don't find this comparison fitting.

▲

CamperBob2 2 days ago | parent [-]

(Shrug) Well, I do. I don't know where you people get the idea that these sorts of things spring forth fully-formed from the brow of Zeus, but it doesn't work that way.

This your first paradigm shift? :-P

▲

roommin 2 days ago | parent [-]

A radio is fundamentally do thing the thing it claims. Artificial intelligence is fundamentally not intelligent.

I'm not saying it is useless tech, but no it's not my first paradigm shift, and that's why I can see the difference.

▲

CamperBob2 2 days ago | parent [-]

How'd you do at the International Math Olympiad this year?

▲

roommin a day ago | parent [-]

You seem to bring much more snark than substance.

▲

CamperBob2 a day ago | parent [-]

Sounds like evasion to me. At some point you'll need to explain exactly why systems that thrash Go grandmasters, make entire forums fail the Turing test, and take gold at international math and programming competitions are "fundamentally not intelligent."

The only argument you (and Toucan, who's been around longer) seem to muster is that the systems aren't perfect. They occasionally say stupid things, need extra handholding, babble nonsense and write buggy code, commit blatant plagiarism, fall into fallacious reasoning traps, and can be fooled with simple tricks... unlike people, presumably.

	▲	roommin 21 hours ago \| parent [-]
		Limited tests in specific settings are always surprising and do speak to it's use, but are not a testament to it being intelligent. I replied because you answered "you're doing it wrong" to a question of it's failures. It seems you dismiss the concerns of the smaller errors or failures without realizing the point being made. If it's "smart" enough to solve take international math medals and beat grand masters at go, but can't truly understand problems and anticipate needs or issues, to me it's not a genuine intelligence and in it's current form never will be. It's not that they are not perfect, is that they have no concept of reality, and it's evident in their failures. Beyond this point I am not interested in trying to convince you.

▲

ToucanLoucan 2 days ago | parent | prev [-]

A radio isn't intelligent and isn't marketed as such. If you're going to sell software you call intelligent, I don't think I'm out of pocket for saying "this feels even dumber than regular software I use."