BOFH: The greatest victory is that which requires no battle Things have taken a bit of a turn for the odd in the IT department. After the initial upset over my replacing the Boss with an increasingly insane AI pseudo-staff member, HR have started to LOVE the idea of AI and are asking for pointers about how to replace other management positions. "But why?" the PFY asks. "Surely HR can …
The article author clearly does not understand how real Human Resources (an oxymoron if ever I heard one!) departments work. Rule No.1 - ANY changes in company staff numbers, from massive expansion to huge redundancies must ALWAYS be accompanied by a large increase in HR staff numbers (to cope with the increased workload, of course). Rule No.2 These staff will be appointed by HR.
Rule No.3 HR staff numbers must NEVER decrease. This leads logically to a business consisting entirely of HR staff and no one else, producing nothing and just consuming resources. I have worked in a couple of large businesses who did seem to be seriously trying to reach this goal, but, oddly, neither of them exist any more.
I consider the case for AI HR to be made and proven.
Why need an actual physical boss to be annoyed and frustrated when the cold hard boot of reality, (BOFH) conflicts with whatever in trend wild eyed idea the Boss comes up.
Why not replace the BOSS with an AI, that does or does not do things as you program it with?
Why not replace everyone with AI?
AI Employeers, AI Employees, AI Customers...
Then what will all of us do lol?
I think we're getting closer to that one scene in the Wally movie, where humans sit in chairs getting fat, and doing not much of anything.
Personally I loathe AI systems, all they are is much better natural language parsing, but with garbage data, and no real intelligence beyond that.
Is AI, a direction we should be going forward in?
Or is this a real indicator that management and business intelligence has been declining for decades?
May the cattleprods always be charging and used!
Personally I loathe AI systems, all they are is much better natural language parsing
A lot of that is due to limitations inherent in the hardware, x86 is not a good fit and GPUs are only a bit better. A completely different processing paradigm is needed if anything approaching AI is to be reached.
Maybe something like this: https://www.theregister.com/2024/05/08/spinncloud_systems_unveils_armbased_neuromorphic/
If you are worried about crap AI, just consider how much worse a fully functional AI could be.
"If you are worried about crap AI, just consider how much worse a fully functional AI could be."
I hate to be a pedant but a fully functional AI would be just as good as the real thing ... so a bit crap, accident-prone, forgetful, but often-enough sufficiently inspiring to make you forgive all that.
Just wait until AI fouls things up like a simple tax return, leading to you getting charged with something serious, and the you getting shuffled off to some hellish purgatory where you're in jail but nobody can help you as things are really fecked up...
I think AI does not belong in decision making positions.
As a tool in helping you to resolve issues, or as an information tool, yes, then that will work.
We still need some humans with common sense at the end of a decision making line, just to check things over.
... still need some humans with common sense at the end of a decision making line ...
Yes, we do.
When did that cease to be a given?
But, as the present AI enabling trend from every TD&H (ie: certified idiot) proves, that option is not necessarily iron-clad, not by a (very) long shot.
Why does the HP/Autonomy affaire instantly come to mind when I read humans with common sense and end of a decision making line in the same sentence?
.
When did that cease to be a given?
Have you ever tried to complain to Google, Meta and their ilk? Computer says no. No humans involved.
And then there was the issue of Uber drivers having their accounts closed with no appeal.
I'm sure there was a story on ElReg concerning the Inland Revenue or DWP, saying that the computer's decision is final with no appeal, but I've just changed my medication and can't find a link.
Everyone should go and watch Terry Gilliam's "Brazil" just to remind themselves of the world we now live in.
... ever tried to complain to Google, Meta and their ilk?
No.
I only use Google (search engine) when I need to sift through and compare with what I can find with other search engines such as DuckDuck.
As to Meta and their ilk, I do not (knowingly/willingly) use any of the so called social media for anything.
ie: metadata harvesting machines.
Learned a very long time ago that whatever convenience these claim to have has a very steep cost.
I also make a point of avoiding any and all shops/pureyors/services that impose the use of WhatsApp or similar for dealing with their customers.
.
With regard to "computer says no" I recently had a 15 year old ebay account nuked within 12 minutes of adding an item for sale, the account had been in use for the whole time buying and selling and had no limitations, the item while fairly high value (£25,000) it was similarly priced to comparable items already for sale.
once the item listed I received an email stating that my account had been permanently suspended and that "The decision is final and can not be appealed."
logging into ebay with the suspended account I found that they had removed all current listings (understandable) but also removed every method of contacting ebay comparing what was shown on my account vs my partners revealed that large chunks of the site were simply missing mainly in the help & support sections.
so in my case it was "Computer says suspended" and "Computer makes sure there is no way for you to complain about it"
"Computer makes sure there is no way for you to complain about it"
There's always the option of a letter to ebay's registered address asserting your rights under GDPR. If that doesn't get a satisfactory response there are legal options. However much they may think otherwise companies running websites actually exist in the real world with real legal obligations, not the fantasy world they believe they're in.
I had an ebay account nuked as well. Not a high value or business account, but one I'd had for years and bought/sold through reliably, no problems or complaints etc.
It was nuked because I logged in from the same IP address that a fraudster had used. The fraudster was an employee of one of my mates and had been offering items for sale that he never delivered, (also some that he did sell and deliver but didn't own - kit he'd been issued with for his job for example, so he's now an ex-employee with a criminal record).
I had visited my mate and connected to his wifi to check on a bid I had running, then a couple of weeks later my account was gone with no warning. It was very difficult to find out why as I hadn't heard of anything similar. It took a while but I finally managed to get in contact only to find out that it was final and there was no appeal (or appeal process or anything).
My mates account was nuked along with his then girlfriends account - she really wasn't happy.
I did consider trying to pursue it but just went 'meh' after a while. Haven't used ebay since.
> When did that cease to be a given?
As soon as the first computer support was introduced - if not earlier.
"The computer / calculator said this, so it must be true" has been a thing for a long time.
Working as a tutor 30 or so years ago, it was already a struggle to get high school students to understand that they needed to have at least a rough idea of what would be a sensible range of results, rather than just blindliy trusting whatever their calculators wound up with.
I had one person write down -E- as the result of a maths question because that's what their calculator showed.
What kind of humans do you think would actually fill those roles?
Do you imagine a wise, thoughtful, fairminded figure who's willing to take time to sift through every decision, every appeal, complaint, whine and whinge, including those generated by bots and politically motivated campaigns, and assess each one fairly and independently on the merits of the case?
Or some weary middle-aged nonentity who's been "promoted" from middle management to this role, who will glance at the rubbish that comes in, spend a few months figuring out how to classify and dispose of it as effortlessly as possible, then never give it another thought?
Or a BOFH who will simply feed their entire intray to another AI, designed chiefly to pick out and disarm those who have the potential to make real trouble?
Humans are only human. If we're relying on them to keep the machines in check, we've already lost.
:-) One small step for aman, one giant leap for Mankind
Humans are only human. If we're relying on them to keep the machines in check, we've already lost. .... veti
Thus is everything already lost to the machines, veti, for humans ARE reliant upon humans to keep them in check.
And ...... if the gospel truth be told, rather than being withheld in order not to terrify and result in revolting natives justifiably disgruntled and furious with current rapidly failing world orders and their systems administrations insisting they have everything under control with the future obeying their every command, your comment that "AI might well be an improvement in many cases. But, I think, not all." is much better shared as AI might well not be an improvement in all cases but certainly in most and that is a great many. for such a much more accurate picture of the upcoming spectacular default situation revision/global reset/alien intervention.
I'm more than just contented and happy to concur with your posit, We're past interesting times, now. We're into downright exciting times. .... for who wouldn't, whenever the above is more true than not, and that which extant status quo leaderships are failing to represent and failing to prevent becoming general knowledge in a common sense appraisal of Earth shattering developments taking place around everyone and everything existing down here on Earth in that space place.
Which is not very bright of them at all, and could even be deadly dangerous to them should they insist on resisting and persisting to not desist in their opposition to that which is clearly a problem for them, and which they have no hope nor any chance of being able to command and control with their plans for things in the future to remain very much the same as they are in the present with them in charge and failing so badly to lead both heroically and epically.
This is a problem that good old-fashioned Market Forces will resolve.
Some companies want to go all in on AI? Let them! Let them replace their middle management, create synthetic experts to replace real ones, implement algorithms to plot their product roadmaps and then follow them...
We'll see how they actually do, against companies run by people.
Of course, there are a lot of badly run companies out there. AI might well be an improvement in many cases. But, I think, not all.
We're past interesting times, now. We're into downright exciting times.
"Why not replace the BOSS with an AI, that does or does not do things as you program it with?"
The key initial is the "I", not the "A". All you need to do is look through Amazon's catalog of reading material, and the suspicion grows that much of it is generated by AI. The same goes for YouTube content. You have to ask yourself about AI, "is it really just doing what you programmed it to do?"
Hmmm...
trolled 1: to fish with a hook and line that you pull through the water - They were trolling the ocean floor- They trolled for fish.2a : to search for or try to get (something) - often: troll for answers/comments/responses- Politicians trolling for votes.
2b: to search through (something) - She loves to troll flea markets looking for bargains.
trawled 1: to pull a large, cone-shaped net through the sea at a deep level behind a special boat in order to catch fish.
2: to search among a large number or many different places in order to find people or information you want: The software is used to trawl for information on the internet - You need to trawl through a lot of data to get results that are valid.
You are both right.
Origin Latin - middle dutch
verb
verb: trawl; 3rd person present: trawls; past tense: trawled; past participle: trawled; gerund or present participle: trawling
1. fish with a trawl net or seine.
"the boats trawled for flounder"
+ catch with a trawl net.
+ sift through as part of a search.
"they trawled through twenty-five-year-old confidential files"
+ drag or trail (something) through water or other liquid.
"she trawled a toe to test the temperature"
noun
noun: trawl; plural noun: trawls; noun: trawl line; plural noun: trawl lines
1. an act of fishing with a trawl net.
"they had caught two trout on the lazy trawl"
+ a large wide-mouthed fishing net dragged by a vessel along the bottom or in the midwater of the sea or a lake.
+ noun: trawl net; plural noun: trawl nets
2. an act of sifting through something as part of a search.
"a constant trawl for information"
3. North American - another term for longline.
In the next episode we will see the BOFH and his sidekick the PFY wrestle control back from their AI overlords and restore order and sanity.
In an alternate universe, the PFY will take a leaf out of the BOFH's grimoire and unleash his own AI (or two) on the company, leading to the unfortunate deployment of the killer bot.
A friend of mine worked in a secret establishment with lots of mathematicians. The best people tend to be slightly strange, lacked social skills, but brilliant at the mathematics. Over 20 years he learned how to recognise the best people to hire (eg wearing odd socks, and a cleanish shirt were good signs).
In the last year or so HR got involved and passed the CV's through an AI system filter before passing "the best candidates" on to my friend.
He said that since then,the people were well dressed, had good social skills, and were the sort of people you put in front of customers. Mathematically they were below what was needed.
My friend complained and was told HR had paid a lot of money for this system, and it was proven to identify people with all round skills who could adapt to do any job - which is what the company needed.
My friend could not get through to them that he didn't want 'all round' people - he wanted the people good at the mathematics.
They could not understand the difference between a group of people with diverse skills - but on average had the skills the company needed; and group of people all with the same average skills.
He retired, but still keeps in contact with some of the people and despairs at "progress"!
"My friend complained and was told HR had paid a lot of money for this system, and it was proven to identify people with all round skills who could adapt to do any job"
Obviously the HR people themselves had failed by their own criteria. They couldn't adapt to the job of recruiting mathematicians.
My current company uses HR AI. I could not have gotten this job, and have already been rejected for promotions because I'm not "pretty" enough to pass the AI interview. I've hit lead tech within a few years at every other company on my job performance, but at this one I'm still sitting at entry level. I've been here long enough that I should have been a senior tech already. So, I've scaled back my work to that expected of an entry level tech. Waiting on The Boss to ask me why I'm not going above and beyond for a promotion.
Yes, I could get another job, but I'm retiring in 2 years or so and can't be bothered. My current job is so easy that it's actually harder to throttle back than it would be to work at my normal capacity, but until I start moving up the ranks properly they're not getting the extra work. Plus, it would be hard to beat working from home on my preferred schedule. I meet my metrics each month, just barely, and precision like this is not that easy.
The "AI" has probably been given rules about promotions for 'junior' employees nearing retirement age.
The "AI" vendors are currently targeting their sales at two groups within any large company - the executive suite and HR. On the basis that these are the people most likely to believe the hype and also believe they are exempting themselves from all liability by using an "AI".
Dr. Susan Blackmore of The Meme Machine regards information passed between machines like this to be a third replicator, alongside biological genes and human-spread memes, which she calls tremes.
Personally I find that view a rather ex-treme meme, but there you go. What does Synthia say?
"When did Simon start speaking American English?
The name of the colour is "grey" in New Zealand English, not "gray". Did someone upset the spill chucker?"
No, they just hit the 'Update' button and it defaulted back to the factory standard of US English, same as Teams always keeps doing at our place, and don't ask me the problem we have trying to get machines to BUILD in UK English.
Quite right. 'AI' is ML. And I see a spooky resemblance in all this to the dotcom bubble that went pop in 2000.
The increasing billions spent on stuff that will definitely be world shattering quite soon - with concomitant huge energy usage.
The cracks are starting to show. Who knew that trawling the internet would lead to breaches of copyright that may lead to millions of lawsuits? Hallucinations? So much data needed that there is now 'cannibalism' of other 'AIs' data? People (including lawyers) depending on 'AI' data that turns out to be totally spurious?
I would think about going to Mars, if I didn't know that I'd either die on the way out or soon after I got there.
Nice one, El Reg/Simon T. Whenever clearly way out in front of both competition and opposition so evidently struggling with and failing to master or take advantage of NEUKlearer HyperRadioProACTive IT opportunities has one certainly leading in all such relative matters by a quite naturally overwhelming almighty otherworldly default.
And of course do/will AIs talk amongst themselves and share both novel and ancient secrets which informs intelligence of the enlightened paths ahead to be pioneered and followed 0Day Exploit Export Hunting for AI Bugs In Search of LOVE ‽
Out there anywhere, IT’s a Brave New More Orderly World Order Toy and Tool readily available virtually everywhere for Work, REST and Play.
Carpe Diem .... and enjoy the AI Way in getting things done better, differently.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
