Phenaut
nixCraft š§IBM slide, 1979 ... š
https://twitter.com/_w0bb1t_/status/1484198603337908226/photo/1
Alyssa Miller š©ļøā 
ā@nixCraft ironically whose one of the companies pushing hard into AI?
Richard Bairwell@nixCraft Holding managers responsible? I note that is one Eastern (especially Japanese) custom which was never adopted by "the Western World".
Nerdeiro 
@nixCraft I could say the same for many C suite execs. If they're not being held accountable by anyone, why let them make decisions ? š¤
Steve@nixCraft
Modern version:
A computer can never be held accountable.
Thus all risky decisions must be delegated to computers.
Gerard Cunningham āļø
James Williams@nixCraft It is itself a management decision to delegate a management decision to a computer, no?
johnaldis@johnaldis @nixCraft Exactly. Itās just a layer of indirection.
@johnaldis @nixCraft But its easy and correct to point at the person who approved the computer. Itās kind of a non issue?
Eddie Roosenmaallen@jamwil @nixCraft
Yes, but itās a second-degree connection, which is easier to escape responsibility for.
If Rick the Accounts Manager round-files 95% of loan applications from Black customers, heās a racist and the company will look bad in the media.
If Rick the Account Manager uses an AI model (based on Rickās history) that rejects 95% of applications from Black customers, well, thatās just science; human managers canāt be held accountable for science. š¤· (/s)
Yes, but itās a second-degree connection, which is easier to escape responsibility for.
If Rick the Accounts Manager round-files 95% of loan applications from Black customers, heās a racist and the company will look bad in the media.
If Rick the Account Manager uses an AI model (based on Rickās history) that rejects 95% of applications from Black customers, well, thatās just science; human managers canāt be held accountable for science. š¤· (/s)
MrRadar@nixCraft As true today as it was 43 years ago. Too bad people so badly want to outsource all their decision making to "artificial intelligence" these days...
Petra DreiskƤmper@nixCraft If only they'd stuck with the plan...
Ministerofimpediments@nixCraft
Computers...making management decisions since 1980 because managers are cowards...
Computers...making management decisions since 1980 because managers are cowards...
Henri Aunin@nixCraft Same should go for AI, shouldn't it?
Sabrina/Brie@nixCraft Corollary: Technology cannot solve societal problems.
The Duality of Xan@nixCraft Counterpoint: A programmer can and sometimes should be held accountable.
Sheila DeBonis@nixCraft So does this mean if the ATS tosses my application, management should retrieve it?
Syd@nixCraft Also works for Elon
Craig Arnold šŗšø šŗš¦@nixCraft
"Computers are dumb and they don't care" - John Hanton, EE professor at Montana State University, 1981.
"Computers are dumb and they don't care" - John Hanton, EE professor at Montana State University, 1981.
Mark Gardner@nixCraft
Same for AI, only it does.
Same for AI, only it does.
Jamie Clark@nixCraft Still accurate. There was never a computing system given power over the rights or affairs of a person, without some other ā¢personā¢ deciding to give it that power. It is that decision maker who bears liability.
The drift in commercial law over the last few decades, where (unlike normal product liability) we refuse to hold software sellers responsible for the damage done by flawed work, does not change this.
When we started rewriting law for electronic commerce, the Americans (UCITA) got this wrong, but the German codes got it right. Unfortunately, most laws have become weaker since then: most Western politicians became utter shills for and utterly dependent on large tech companies, and let liability, responsibility and competition law go limp.
HT @rigo
The drift in commercial law over the last few decades, where (unlike normal product liability) we refuse to hold software sellers responsible for the damage done by flawed work, does not change this.
When we started rewriting law for electronic commerce, the Americans (UCITA) got this wrong, but the German codes got it right. Unfortunately, most laws have become weaker since then: most Western politicians became utter shills for and utterly dependent on large tech companies, and let liability, responsibility and competition law go limp.
HT @rigo
James š¦ #FBPE 
Max Leibman has moved!@nixCraft Many are confused on this point, but to date, a computer has never made a decision of any sort.
The responsibility for every computerized ādecisionā lies with one or more humans. Those who design, code, train, deploy, and decide to act on the outputs of such systems are 100% responsible for the outcomes.
Bots arenāt culpable. People are. There may greater or lesser layers of convolution and obfuscation between the humans and the apparent decision, but humans are always responsible.
Karl@nixCraft we are not a very smart species.
mike@nixCraft I have thoughts about this. But first I must go praise the dawning singularity.
Meredith@nixCraft objection- computerās fault when it broke and code was suck and didnāt work!
Jon Hatch 
@nixCraft sometimes I just let my roomba sit under the couch for a while.
"You made this decision, now you face the consequences"
Scott Williams š§@nixCraft This seems to imply that those who make management decisions are held accountable.
josh
Astra Kernel @nixCraft but it might be more intelligent in decision making than most people.
Lev Petrovitch@nixCraft
That's what computer industry had to say. The rest of the capitalists went
"Holy cow, you mean we can have a computer make the decisions we want and nobody will be held accountable?"
Thomas la Cour Jansen@nixCraft time to release the new OpenAI management programā¦
MWDelaney@nixCraft this is how you avoid a Butlerian Jihad. Thumbs up.
@nixCraft neither can a corporation
Paul Bradish@nixCraft fair point, lol.
Earth Notes@nixCraft There's a lot of (buggy) spreadsheets out there that managers have used to make decisions with only the thinnest smear of human agency on top...