How Much AI Do We Need?

Question

How Much AI Do We Need?

Steve Jones - SSC Editor

SSC Guru

Points: 736212
More actions
April 20, 2020 at 12:00 am

#3742902

Comments posted to this topic are about the item How Much AI Do We Need?

Viewing 9 posts - 1 through 8 (of 8 total)

You must be logged in to reply to this topic. Login to reply

skeleton567 SSCertifiable Points: 5868 More actions · Answer 1

Depending on AI is like listening to the news. It is more opinion than fact. It is by its very nature biased by the beliefs and desires of those who create it. Why should I trust others handling of the facts more than I trust my own?

Is it not amazing that people are getting all bent out of shape over the current limitations on 'rights' due to the virus thing but allow software to control their lives?

Rick
Disaster Recovery = Backup ( Backup ( Your Backup ) )

Jeff Moden SSC Guru Points: 1004340 More actions · Answer 2

skeleton567 wrote:

Depending on AI is like listening to the news. It is more opinion than fact. It is by its very nature biased by the beliefs and desires of those who create it. Why should I trust others handling of the facts more than I trust my own?
Is it not amazing that people are getting all bent out of shape over the current limitations on 'rights' due to the virus thing but allow software to control their lives?

Man, I couldn't have said it better. +1,000,000!

To answer the question of "How much AI do we need"? My answer would be 1) Most of it isn't actually "AI" and 2) whatever you want to call it, we need a whole lot less of it.

--Jeff Moden

RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
First step towards the paradigm shift of writing Set Based code:
________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

Change is inevitable... Change for the better is not.

Helpful Links:
How to post code problems
How to Post Performance Problems
Create a Tally Function (fnTally)

Doctor Who 2 SSCrazy Eights Points: 8194 More actions · Answer 3

I hear you, Steve. There's no one-size-fits-all type of AI. Your analogy of cooking and baking is excellent. I'm a terrible cook, so if I were left on my own I'd go for all of those smart devices. My wife is a wonderful cook, for her those smart devices just get in the way.

Probably its the same with AI in that some who are less technically savvy would want all the bells and whistles. Whereas others who are much more skilled, technically, would go for whatever would supplement their own skillset.

Rod

Jeff Moden SSC Guru Points: 1004340 More actions · Answer 4

Doctor Who 2 wrote:

I hear you, Steve. There's no one-size-fits-all type of AI. Your analogy of cooking and baking is excellent. I'm a terrible cook, so if I were left on my own I'd go for all of those smart devices. My wife is a wonderful cook, for her those smart devices just get in the way.
Probably its the same with AI in that some who are less technically savvy would want all the bells and whistles. Whereas others who are much more skilled, technically, would go for whatever would supplement their own skillset.

I do agree with that but what I've found, especially in SQL Server and automobiles, is that you sometimes aren't given a choice.

For example, the "AI" (I hate to call it that) says that every once in a while, it's going to flash on the "'tube" display that taking your eyes off the road too long to look at that very display is a bad idea. It then requires you to press a virtual button on the display to get your normal display back even if you're traveling at 70 miles per hour! That same display is also so bright with the backup camera when you're in reverse that you can't actually see out the side window to see the bloody mirrors... and there's no setting to turn it down! I realize that's not "AI" but many of the same kind of people that design that kind of nonsense functionality also design supposed "AI". I also totally disagree with what people are calling "AI".

To wit, "bells and whistles" can be great but only if you can silence them or train them to work better. I also believe that every such device needs to have an "STFU" override button and a voice command to match. 😀

--Jeff Moden

RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
First step towards the paradigm shift of writing Set Based code:
________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

Change is inevitable... Change for the better is not.

Helpful Links:
How to post code problems
How to Post Performance Problems
Create a Tally Function (fnTally)

xsevensinzx One Orange Chip Points: 25560 More actions · Answer 5

So many confusing ML with AI or AI with ML.

I often point to the example of Google.com where it's the ML that is behind the scenes mining the keywords, but it's the AI telling the ML what keywords to mine and display as the suggestion when you type.

Jeff Moden SSC Guru Points: 1004340 More actions · Answer 6

To me, most machines/computers/whatever aren't "intelligent" nor are they able to "learn". At least, not yet. Don't get me wrong... there's some very, very clever programing being done out there that give people such an impression but, so far, I wouldn't classify most of it as "intelligent" nor "learning".

--Jeff Moden

RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
First step towards the paradigm shift of writing Set Based code:
________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

Change is inevitable... Change for the better is not.

Helpful Links:
How to post code problems
How to Post Performance Problems
Create a Tally Function (fnTally)

xsevensinzx One Orange Chip Points: 25560 More actions · Answer 7

Jeff Moden wrote:

To me, most machines/computers/whatever aren't "intelligent" nor are they able to "learn". At least, not yet. Don't get me wrong... there's some very, very clever programing being done out there that give people such an impression but, so far, I wouldn't classify most of it as "intelligent" nor "learning".

That's not entirely accurate. I feel you are discrediting these because they require some human intervention. I must stress, the act of learning is the act of acquiring knowledge through study. Machine learning is absolutely acquiring knowledge through study, but it has to be taught how to study first too. That's ideally where the human intervention comes into play and that's no different than humans being taught by a teacher either.

We are not born to just know how to read, what the color of red is, or what is right or wrong. We are taught this and so will the machines we use. Thus, as they are taught these facts and as they analyze patterns of data, they will absolutely have some bias on what to do and what not to do in many cases. However, there is the fact that new patterns in the data will be discovered much like discovering a new color you weren't taught by your art teacher, and that ideally where a big portion of the learning and even AI will come into play in how the machine reacts to that new piece of knowledge.

But, make no mistake, the act of acquiring that piece of knowledge is the act of learning. How the machine acts on that piece of learning is where the AI really comes into play. And we are just scratching the surface of how we design the machine to ideally think for themselves or make rational decisions without as much human intervention or bias.

Jeff Moden SSC Guru Points: 1004340 More actions · Answer 8

I guess we'll have to agree to disagree on most of that.

--Jeff Moden

RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
First step towards the paradigm shift of writing Set Based code:
________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

Change is inevitable... Change for the better is not.

Helpful Links:
How to post code problems
How to Post Performance Problems
Create a Tally Function (fnTally)