81 comments
@berkes @i0null If the computer misbehaves you can't punish it. And most of the time it misbehaves because somebody designed a component (or many) wrong. Therefore you shouldn't rely on the computer to make decisions that can't be reversed or appealed by humans, because there needs to be a human higher in the hierarchy than the computer who can fix things. @Natanael_L @i0null seriously: why can you not "punish the computer"? We can punish a car manufacturer when they forgot to put in brakes. We can punish an app developer when their app is harming us. We can punish a cloud provider when they leak data. And so on. I feel it's far more involved than you state. What if I don't use "a computer" but instead a lambda on AWS? It would make sense the software developer is responsible when it makes a Wong decision? @berkes @i0null you have the whole chain from collecting requirements to implementation to the operator of the system, the responsibility lies with whoever has contributed to the misbehavior (or opted not to fix it). Whoever made the decision that software or hardware which they should have known wasn't ready should be put into production. @Natanael_L @i0null In that case "the computer" being a system, can be held responsible, no? @berkes @Natanael_L @i0null One can't 'punish a car manufacturer'.. one can add to the expense of manufacturing but that is just another of many that businesses deal with in every transaction. @Natanael_L @berkes @i0null Only humans higher in the hierarchy refuse to oversee the decisions algorithms make. See almost any ban appeal in social media. Or complaint about bank handling charges. @llywrch @Natanael_L @i0null Certainly. But a counterexample were software developers and managers that went to jail in the VW scandal. So there is accountability for those writing software. Not much and not always. But not "never" either. @berkes @i0null We've already seen this in action for a while now: "the algorithm messed up," "the algorithm did something unexpected," "the algorithm shouldn't have done that," the algorithm this, the algorithm that. It's a thought-terminating cliche and, unfortunately, it generally works on people. @i0null please include sourcing if you can. Unfortunately itβs a little vague in this case, as the original was destroyed in a flood: https://twitter.com/bumblebike/status/1468346709994582020 @glyph @i0null Text including "The computer mandate" is visible through the paper. Considering the user took a photo of the hole-punched document on their knees and says they worked at IBM and in UX, it's reasonable to believe that this is an internal training document rather than an elaborate hoax. @glyph @i0null They uploaded this page too. https://twitter.com/bumblebike/status/1160292780574298112 @mattjhodgkinson @glyph @i0null i really wish we had the full thing, there must be other bangers like this in it. @i0null it's been a long time since any IT industry executive could be held accountable.. * "The Analytical Engine has no pretensions whatever to originate anything. It can do whatever we know how to order it to perform." Ada Lovelace @i0null A BILLIONAIRE CAN NEVER BE HELD ACCOUNTABLE. THEREFORE A BILLIONAIRE MUST NEVER MAKE A MANAGEMENT DECISION. @i0null @mastodonmigration Counterpoint: human managers arenβt really held responsible for their decisions either, so not convinced thatβs a valid blocker anymore @dogzilla @mastodonmigration if one substitutes the word βcomputerβ with βsystemβ or a βcollection of prescribed rulesβ then semantically the point still applies. @i0null This is exactly the reason why AI decision making is so popular (e.g. in fraud detection and insurance). Deciding with algoritmes which are heavily biased without management being accountable. βThe machine says noβ. @i0null@infosec.exchange the problem with this slide is it seems to imply that human management can be held accountable. @i0null Corrolary: humans who cannot be held accountable must also never make management decisions. @i0null was the next slide "UNLESS THE DECISION IS REALLY UNCOMFORTABLE, LIKE DENYING TREATMENT, IN WHICH CASE THE COMPUTER IS G2G"? Replace "computer" with 'billionaire' and you get a fairly approximate explanation of the world we're unfortunately living in right now. someone has probably already said this , but I will bet dollars to donuts that in 1979, the very large IBM salesforce was pushing computers as a way to make management decisions and probably selling a lot of 360s on that basis Replace the word computer with Republican and viola! You have the sad state of this shithole country today. Uh, think you are missing something, management is seldom or never held accountable, this is the system. Why unions exist for labor. @i0null the whole history of politics and bureaucracy is looking for and making legal ways to dillute accountability to the point of almost absolute anonimity and zero responsibility of a given functionary. the quote above basically is a solution to that Β«problemΒ». @i0null This was clearly written by someone in Management because only theyβd be living under the romantic idea that Management is actually held responsible for anything and is punished accordingly. The rest of us gave up on that idea a long time ago. @i0null So, this is why I can't advance? π So willing to hold me accountable for Bonnie though. Even with flimsy evidence. @i0null When I worked at IBM, the people working there for decades would lament the loss of the company who used to take care of their people. It actually reminds me of the feeling of loss that the people at Google have been feeling these past few years. @i0null Therefore a human must never make a management decision. Trust in Friend Computer. @i0null @i0null wait is there such a thing as accountability in management decisions? When was that invented? I thought we were still blaming interns My favorite statement from IBM manuals is still current: "MVS bases its decisions and activities on the assumption that time is progressing forward at a constant rate." https://www.ibm.com/docs/en/zos/3.1.0?topic=parameter-setting-tod-clock That's a pretty good assumption, as long as you steer clear of black holes. "That's why at Torment Nexus all decisions affecting people's lives are made at the retail level, closest to the customer." @i0null The premise of this was largely nullified when Citizens United decided that corporations have citizens rights. The idea that an entity needs accountability went out the window right there, to the extent that it ever even existed. @i0null @RiaResists Senior management is rarely held accountable! Even when their decisions result in massive death counts. The Sacklers should be on death row. My skeptic thinks it could be a sales pitch... "Managers, buy this computer. It will not eliminate YOUR job." A corporation can never be held accountable. Therefore a corporation must never make a management decision. @i0null IMHO the company developing the software should be held accountable. Apply this to AI and I hardly think anyone would even dare releasing AI solutions to the public just yet π @i0null@infosec.exchange A cat can never be held accountable therefore cats must make ALL management decisions! |
@i0null I've spent way too long pondering this quote lately.
I think the problem sits in the premise. Is it true that "a computer can never be held accountable"?
And in that, what is "a computer", is it just the physical metal and silicium, the entire product, the entity renting it out, the entity using it, or owning it?
And is it truly "never"? As in: is this something dictated by fysics or the Order of Things, or can this change?
Like I said. Way too long.
@i0null I've spent way too long pondering this quote lately.
I think the problem sits in the premise. Is it true that "a computer can never be held accountable"?
And in that, what is "a computer", is it just the physical metal and silicium, the entire product, the entity renting it out, the entity using it, or owning it?
And is it truly "never"? As in: is this something dictated by fysics or the Order of Things, or can this change?