SQL Clone
SQLServerCentral is supported by Redgate
 
Log in  ::  Register  ::  Not logged in
 
 
 


Badly Trained AI


Badly Trained AI

Author
Message
Steve Jones
Steve Jones
SSC Guru
SSC Guru (628K reputation)SSC Guru (628K reputation)SSC Guru (628K reputation)SSC Guru (628K reputation)SSC Guru (628K reputation)SSC Guru (628K reputation)SSC Guru (628K reputation)SSC Guru (628K reputation)

Group: Administrators
Points: 628503 Visits: 21336
Comments posted to this topic are about the item Badly Trained AI

Follow me on Twitter: @way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com
Dave Poole
Dave Poole
SSC Guru
SSC Guru (63K reputation)SSC Guru (63K reputation)SSC Guru (63K reputation)SSC Guru (63K reputation)SSC Guru (63K reputation)SSC Guru (63K reputation)SSC Guru (63K reputation)SSC Guru (63K reputation)

Group: General Forum Members
Points: 63759 Visits: 4050
I worked on a personalisation system that affected the sort order of products returned from a search so as to place the products in order of the customer's propensity to buy.

After the initial excitement had worn off we found that the identified preference became stronger and stronger. Not because the customer's ACTUAL preference was any stronger but because they were only seeing products based on past preferences (unless they started paginating through results which is something very few customers did).

We identified the need to introduce a random element into the mix. Using machine learning techniques to produce an algorithm and dataset identifying likely customer behaviour based on past behaviour was relatively easy. Doing the same thing to identify the optimum ratio of random elements was a problem we didn't crack.

One of the dangers we identified was that initial results were so good as to lull an organisation into a false sense of security. The self fulfilling prophecy element would lead to customer boredom and associating your brand with a very limited range of products even if in reality the brand offers a very wide range. It is easier and cheaper to retain a customer than to acquire/reacquire them.

Another aspect that is missing is the skill in salesmanship. If ML deduces that women tend to put on weight as they age then you have to be damn careful how you market clothes to address a fact that the customer may not be comfortable with.

LinkedIn Profile
www.simple-talk.com
robert.sterbal 56890
robert.sterbal 56890
SSCrazy Eights
SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)

Group: General Forum Members
Points: 9557 Visits: 1722
I prefer automation to intelligence. Baselines to predictive analytics.
Eric M Russell
Eric M Russell
SSC Guru
SSC Guru (113K reputation)SSC Guru (113K reputation)SSC Guru (113K reputation)SSC Guru (113K reputation)SSC Guru (113K reputation)SSC Guru (113K reputation)SSC Guru (113K reputation)SSC Guru (113K reputation)

Group: General Forum Members
Points: 113241 Visits: 15180
David.Poole - Thursday, January 10, 2019 2:08 AM
I worked on a personalisation system that affected the sort order of products returned from a search so as to place the products in order of the customer's propensity to buy.
After the initial excitement had worn off we found that the identified preference became stronger and stronger. Not because the customer's ACTUAL preference was any stronger but because they were only seeing products based on past preferences (unless they started paginating through results which is something very few customers did).
...

When online shopping, I typically just enter a couple of keywords and then immediately sort the returned search results by price from lowest to highest.
Satisfied



"The universe is complicated and for the most part beyond your control, but your life is only as complicated as you choose it to be."
Jonathan AC Roberts
Jonathan AC Roberts
SSChampion
SSChampion (11K reputation)SSChampion (11K reputation)SSChampion (11K reputation)SSChampion (11K reputation)SSChampion (11K reputation)SSChampion (11K reputation)SSChampion (11K reputation)SSChampion (11K reputation)

Group: General Forum Members
Points: 11992 Visits: 5288
One thing I don't understand about biased AI systems is that how can AI system can be biased if you don't give it the information that could make it biased? For example, if you don't give the algorithm the sex and race of the person how could it come to a biased conclusion about the person?
Jeff Moden
Jeff Moden
SSC Guru
SSC Guru (907K reputation)SSC Guru (907K reputation)SSC Guru (907K reputation)SSC Guru (907K reputation)SSC Guru (907K reputation)SSC Guru (907K reputation)SSC Guru (907K reputation)SSC Guru (907K reputation)

Group: General Forum Members
Points: 907015 Visits: 48688
Heh... AI... machine learning... When it comes to advertising on the net, someone is hitting the crack pipe pretty hard. I leased a new truck a year ago and immediately got slammed with advertising and emails try to entice me to... buy a new truck. I'm still getting that kind of advertising although at a slightly less volume. It's just stupid and incredible waste of advertising dollars. I know how that advertising is paid for because I used to work for a company that did "Double-click.net" processing for "spotlight pixels" in every graphic that showed up. Huge waste and stupid waste of money the way I've been hit. No wonder why stuff costs so much nowadays.

--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.
If you think its expensive to hire a professional to do the job, wait until you hire an amateur. -- Red Adair

When you put the right degree of spin on it, the number 318 is also a glyph that describes the nature of a DBAs job. Wink

Helpful Links:
How to post code problems
How to post performance problems
Forum FAQs
DinoRS
DinoRS
Ten Centuries
Ten Centuries (1.3K reputation)Ten Centuries (1.3K reputation)Ten Centuries (1.3K reputation)Ten Centuries (1.3K reputation)Ten Centuries (1.3K reputation)Ten Centuries (1.3K reputation)Ten Centuries (1.3K reputation)Ten Centuries (1.3K reputation)

Group: General Forum Members
Points: 1292 Visits: 616
Well not just Amazon's attempt at AI has been biased, Google and Microsoft AIs have been found to be racist at times.

I would love to have a simple AI / ML approach to environment Monitoring, there are paid solutions out there which offer "predictive failures" and things like that but I'd be happy if I could use one SQL Server to highlight all Queries from all servers which degrade in performance or show other fuzziness happening.

There is also at least 1 Insurance Company in China which does customer website chat sessions based on AI to answer questions and inquiries, not sure how satisfied they are, tho.
robert.sterbal 56890
robert.sterbal 56890
SSCrazy Eights
SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)SSCrazy Eights (9.6K reputation)

Group: General Forum Members
Points: 9557 Visits: 1722
I actually came up with a pretty good test for transcription AI. Use the recordings of the public domain works on Librivox.org and compare the transcript to the public domain work.

It solves the volume, comparison and usage issues.

Tests are hard to make vendors take.
xsevensinzx
xsevensinzx
SSC-Insane
SSC-Insane (22K reputation)SSC-Insane (22K reputation)SSC-Insane (22K reputation)SSC-Insane (22K reputation)SSC-Insane (22K reputation)SSC-Insane (22K reputation)SSC-Insane (22K reputation)SSC-Insane (22K reputation)

Group: General Forum Members
Points: 22815 Visits: 6089
Jeff Moden - Thursday, January 10, 2019 8:19 AM
Heh... AI... machine learning... When it comes to advertising on the net, someone is hitting the crack pipe pretty hard. I leased a new truck a year ago and immediately got slammed with advertising and emails try to entice me to... buy a new truck. I'm still getting that kind of advertising although at a slightly less volume. It's just stupid and incredible waste of advertising dollars. I know how that advertising is paid for because I used to work for a company that did "Double-click.net" processing for "spotlight pixels" in every graphic that showed up. Huge waste and stupid waste of money the way I've been hit. No wonder why stuff costs so much nowadays.

I work in advertising myself. Got to remember that when you think about all of the people in the world or all of the people in your country, you are marked as someone you MAY have an interest to buy a car compared to the hundreds of millions or billions around the world where they have no idea if they can either drive. So, while you may think it's a waste, it's still a pretty big lead on a potential sale for someone from a statistical standpoint.

But you know, the leads can go cold! I guess that what separates the good advertisers from the bad ones.

jay-h
jay-h
SSCoach
SSCoach (16K reputation)SSCoach (16K reputation)SSCoach (16K reputation)SSCoach (16K reputation)SSCoach (16K reputation)SSCoach (16K reputation)SSCoach (16K reputation)SSCoach (16K reputation)

Group: General Forum Members
Points: 16434 Visits: 2818
The problem with AI is that it sometimes gives answers based on data rather than the answers desired.

Of course ff the data is indeed bad the results are bad, but another part of the problem is that we live currently in a world dictated by reality, which doesn't always statistically mesh with ideological ideals. AI has no inherent bias, so the next best thing is to blame the data. Or discrimination.

...

-- FORTRAN manual for Xerox Computers --
Go


Permissions

You can't post new topics.
You can't post topic replies.
You can't post new polls.
You can't post replies to polls.
You can't edit your own topics.
You can't delete your own topics.
You can't edit other topics.
You can't delete other topics.
You can't edit your own posts.
You can't edit other posts.
You can't delete your own posts.
You can't delete other posts.
You can't post events.
You can't edit your own events.
You can't edit other events.
You can't delete your own events.
You can't delete other events.
You can't send private messages.
You can't send emails.
You can read topics.
You can't vote in polls.
You can't upload attachments.
You can download attachments.
You can't post HTML code.
You can't edit HTML code.
You can't post IFCode.
You can't post JavaScript.
You can post emoticons.
You can't post or upload images.

Select a forum









































































































































































SQLServerCentral


Search