[HN Gopher] When the algorithm is your boss
___________________________________________________________________
When the algorithm is your boss
Author : jrepinc
Score : 50 points
Date : 2022-01-30 19:29 UTC (3 hours ago)
(HTM) web link (tribunemag.co.uk)
(TXT) w3m dump (tribunemag.co.uk)
| RcouF1uZ4gsC wrote:
| This phrasing is already conceding half the battle.
|
| A computer or "algorithm" doesn't decide goals and values on its
| own.
|
| The computer executes an algorithm that efficiently implements
| the values of decisions of the people in charge of developing and
| deploying the software (and their bosses).
|
| Businesses have done a great job of laundering controversial and
| sometimes illegal decisions through "algorithms" as though they
| were some kind of independent entity.
| Buttons840 wrote:
| The "Uno Reverse Card" for this situation is to ask why we need
| executives and management when the algorithm does everything.
| The answer is, as you say, the algorithm isn't actually doing
| anything really, just what its owners would have done
| themselves.
| jt2190 wrote:
| I agree. The interests of managers are often in direct
| conflict with the interests of owners, as the managers try to
| carve out a larger and larger piece of the business for
| themselves. These algorithms will reduce the need for
| managers, which is what owners want.
| ahelwer wrote:
| I don't think that's quite true. It's like the difference
| between surveilling someone by tailing them and surveilling
| them with an automated network of cameras equipped with
| facial recognition. Scale matters. Much work (especially work
| we call unskilled) is rendered tolerable by the small human
| moments you have with your coworkers, a conversation out back
| when taking out the trash or whatever. Management sees any
| such thing as "time theft" or a similarly dystopian term and
| would like to eradicate it, but they cannot be everywhere at
| once. An algorithm _can_ be everywhere at once. So the
| conversation "maybe people should not actually have to be
| doing productive things the entire time they are being paid"
| never really had to be had before, because people would just
| do things to make their job tolerable. But now that such
| simple pleasures can actually be eradicated the conversation
| does have to be had.
|
| There are other domains where analogous conversations should
| be had. Many people agree with their laws in the general and
| even specific senses, but almost nobody would want to live in
| a society where all the laws are enforced perfectly 100% of
| the time. Occasional lawbreaking is necessary for society to
| function. Anybody who objects to this should really think
| through what they're wishing for.
| lupire wrote:
| That's a different topic -- how much power one person
| should be allowed to have. It's still the person's power.
| zwkrt wrote:
| I agree mostly but i think that saying "the algorithm is my
| boss" really does help the discussion along. I had a friend
| ("S") who on hard times sighed up to deliver through uber eats.
| Their entire tenure included no human interaction. S worked for
| about 3 weeks, got some bad reviews for delivering food cold,
| and were "fired" all without ever talking to anyone or seeing
| anyones face. They would ironically (but truthfully!) claim
| that their phone was the worst boss they ever had.
|
| Its true that somewhere behind the scenes is a set of people
| making the algorithm and a set of people implementing it, but
| for the end user/worker there is no human, no realistic
| arbitration, no negotiation, no favors, no coworkers, no
| smiles, no understanding, and no recourse. It reminds me of
| yesterday's "Did I just lose $500k?" post. When S got fired it
| may as well have been mandated by god.
|
| ADDENDUM: I don't want to toot our horn too much, but also
| remember that as a member of this forum you are probably an
| order of magnitude better at navigating complex institutions,
| understanding business objectives, understanding the "intent"
| if a piece of algorithmic technology, and advocating for
| yourself than the average person who picks up gig work. To many
| people it is literally magic, and the further your boss is
| toward being magical the closer you are to being a slave to
| them. And I don't think I am being melodramatic with my choice
| of words.
| lupire wrote:
| The phone wasn't their boss. Uber's bilionaire management was
| their boss. Phone was the scapegoat.
| rambambram wrote:
| > ... remember that as a member of this forum you are
| probably an order of magnitude better at navigating complex
| institutions, understanding business objectives,
| understanding the "intent" if a piece of algorithmic
| technology, and advocating for yourself than the average
| person who picks up gig work. To many people it is literally
| magic, and the further your boss is toward being magical the
| closer you are to being a slave to them.
|
| So true. Not everybody is going to see through the smoke and
| mirrors.
| Animats wrote:
| Machines should think. People should work.
|
| This is so much more cost-effective than slavery.
| jpalomaki wrote:
| On the other hand, with algorithms you can actually study how the
| decisions are made. You can investigate what was given as input
| and how it affected the results.
|
| With complicated algorithms it may not be obvious, but then you
| can try to poke the black box with thousands of sample cases to
| see how it reacts.
|
| Compare this with humans, who often don't even know themselves
| what their decisions are really based on.
| dehrmann wrote:
| Yeah, it's easier to investigate a racist algorithm than a
| racist manager. The article makes the following point, but I'm
| not sure if it's necessarily worse than the status quo:
|
| > All the while, the potential for discrimination based on
| race, gender, or disability, especially in hiring and firing
| software, remains obscured behind the black box of AI.
| advisedwang wrote:
| Who is the "you" here? Workers certainly do not get to probe
| the internals of the algorithm that fired them. For liability
| reasons I'm sure human resource management systems are not
| exposed to researchers. Maybe a court could but in practice I
| don't think that's happening and the bar to get that kind of
| review is incredibly high.
|
| Only the company itself is actually in practice going to be
| testing their system but they only do so with a view to their
| own interests.
| giantg2 wrote:
| "On the other hand, with algorithms you can actually study how
| the decisions are made. You can investigate what was given as
| input and how it affected the results."
|
| You'd have to have access to it. This is highly unlikely. If
| testing as a black box, could they identify a test and change
| its behavior, like VW emissions tests?
| the_snooze wrote:
| Machines don't have shame, reputation, values, empathy. There's
| little hope of self-correction. An algorithm will never ask
| "Are we the baddies?"
| verisimi wrote:
| This is the planned future for us all, unfortunately. Technocracy
| demands that scientific experts evaluate our reality and then
| determine what we are and are not allowed to do.
|
| Too much carbon/water/energy use = restraints on your usage. (No
| restraints if you can pay though.) This is what smart meters,
| smart phones, smart cities are going to facilitate. (Smart = Spy)
|
| If I agreed with their evaluations and was able to choose to opt
| in, I might even consider this. But this is not the plan. What
| will really happen is that the political and billionaire classes
| will find that it is right for them to assume the role of
| determining the technocratic goals that the AI driven system
| should achieve.
|
| We will find that we have a tyrant computer to engage with. There
| won't be a friendly face to help you by bending the rules in some
| way. It will be like when you ring up your local governance
| office to sort something out... but worse. We are not going to be
| stepping into techno-utopia, that's for sure.
|
| A bio-medical-security-id seems to be a required first step.
| Gys wrote:
| > What will really happen is that the political and billionaire
| classes will find that it is right for them to assume the role
| of determining the technocratic goals
|
| 'They' have power and money because 'we' give it to them. Not
| because they take it from us.
| verisimi wrote:
| Yes. Partly we give it to them, because they do not present
| their case in a straightforward way. I think there is a lot
| of social engineering that occurs, that just happens to align
| and progress the pre-planned agenda. So, imo, we give our
| money and power over, as we are deceived.
| zackmorris wrote:
| This was supposed to be a warning, not an instruction manual, but
| here it is again:
|
| https://marshallbrain.com/manna
| not2b wrote:
| Note that this is a British article, and where it uses the word
| "liberal" American should read it as somewhere between
| "libertarian" and "neoliberal".
| rambambram wrote:
| Triangulation (the narcissistic one) at it's best. Make a machine
| - or software function in this case - the messenger. As an evil
| c#nt you get a) plausible deniability, b) smoke and mirrors, c)
| gas lighting, and d) a scape goat in one. Oh, and your algorithms
| become your 'flying monkeys' as well.
|
| For personal reasons I had to dive into NPD, BPD and psychopaths
| (all cluster B personality disorders) the last years. Once I
| finally understood that sh1t, a whole lot of other stuff made
| sense. What big tech is doing here is using these same principles
| on a world wide scale. I'm not saying this is new, because the
| Romans with their 'divide and conquer' did basically the same.
|
| The solutions to counteract this as an individual are also the
| same as found in the psychology books: go 'no contact', or at
| least 'gray rock'. If you are capable, that is. There's going to
| be a lot of casualties, and I feel for the workers there (or
| should I say 'victims').
| heisenbit wrote:
| Yup. Impulsive decision making, lack of empathy, lack of
| ability to take different points of view and black&white
| thinking.
| rambambram wrote:
| Lack of empathy? Definitely. Impulsive decision making? I'm
| not so sure.
|
| What you describe sounds pretty BPD, especially with the
| black/white thinking (although I think hot/cold is clearer
| wording). I was more thinking along sinister lines here.
|
| And indeed, zero regard for different points of view. Which
| is very logical when one finally understands that: if you are
| - in your own mind - the god of the universe and beyond,
| every other view is by definition beneath you.
___________________________________________________________________
(page generated 2022-01-30 23:00 UTC)