IBM to pause hiring in plan to replace 7,800 jobs with AI
finance.yahoo.com
finance.yahoo.com
Expect to see more of this from zombie companies in the future, as long as they spend $20/mo on a ChatGPT subscription somewhere they are probably not violating SEC disclosures either.
This hasn't been the case with the ongoing tech layoffs though? Investors for the most part congratulate these companies for diligently cutting costs in preparation of a recession.
If a company that people are expecting to be losing relevance does a layoff, it’s still not a good sign.
I guess they tossed all the services/consulting into Kyndryl and everything else got to stay blue.
The missing pieces (in my subjective opinion) are: (1) happy AND talented employees and (2) a rapid execution culture.
I was making the point that they have a solid platform available to build almost anything the FAANG companies have. Of course, they will have to build them. For that, they need to hire great engineers and they need to retain and keep them happy.
(Quick note: Traction is a whole different ballgame, but there are companies who gained initial traction through the customers of their business clients. In IBM's case, that is almost every major government agency and Fortune 500 businesses.)
Most of the business consulting, including outsourcing software engineering, is still with IBM.
I know they got rid of OS's and laptops a while back; I assume they don't make or service mainframes/minicomputers anymore. Do they do any software/hardware development, or is it all consulting?
Still plenty of hardware and software stuff done via IBM with POWER systems, AIX, etc. They were an early investor in Red Hat and now own it entirely.
I have a few contacts with them and mostly they just rested and (not) vested.
Most companies don't have the capitalization, runway or people who are so badly broken as to go through the bidding process for something like that.
So, IBM using this opportunity to cut down their own org tree a bit more and "replace it with AI" (or simply drop some people) is not unexpected. Because consulting related to doing exactly that is probably what they'll be selling a lot for the next few years. Basically, as world + dog is trying to figure out how to fit AI into their businesses, consulting related to that is going to be booming. And IBM is a consulting company. A lot of their previous efforts are actually powering the backoffice of course. So now that AI is going to replace all of that, that basically means there is going to be a lot of consulting happening as to how that would work. So that's a good business opportunity.
Consulting is less than a third of their revenue and regardless of how you feel about consulting, it seems kind of idiotic to claim that this "barely makes sense at all at this point". Are the only companies that make sense the ones that work on fad bullshit that makes the front page of HN?
The hot new thing at IBM was "global services" consulting. Since I left they've sold their chip business, laptop business and really doubled down on consulting..
Did you get to work with Edgar Codd or do you know someone who might know someone who worked with him?
That said, I've hated the handful of IBM software products I've needed to use.
Another anecdote: I've never met a fellow developer that spoke highly of IBM. That includes the folks who used to work there.
I think it's remarkable how rare it is to meet someone that speaks highly of the company. Companies don't usually achieve such reputation without reason.
Worked with a buncha folks in the NC Research Triangle area, all former IBM. None had good things to say about the engineering culture there. Most did long stays regardless -- it's that kinda of org -- talking like 10+ years, but jumped to other companies like Red Hat (ha!), Clorox, Splunk, and Uni research after the last big shakeup.
A better model was that Rometty was not good at choosing what to fund and which bets to make. kinda a failed VC fund.
Lookup the saga of HP's memristor followed by "The Machine" for a similar story.
Then I guess we can throw out Xerox Parc's NLS because they never commercialized "windows, hypertext, graphics, video conferencing, the computer mouse, word processing, dynamic file linking, revision control, and a collaborative real-time editor." They were commercialized by others later on.
https://en.wikipedia.org/wiki/The_Mother_of_All_Demos
We can also throw out most psych, social science, archeology, and history R&D, too.
And the laser, which took a few decades to find an application.
Fun fact: funded by ARPA
Archeology / history, on the other hand, definitely have some pretty awesome finds.
Yes.
It attracts playful wizards. If you get enough of those no amount of zombies can beat them.
They could have built it over time to offer real assistance to knowledge workers with a real revenue stream.
But that would have required vision and foresight.
https://www.nytimes.com/2021/07/16/technology/what-happened-...
However everyone is still pretty far off from that.
> regardless of how you feel about consulting
It's a matter of how well they do it. They're not known for creating customer value.
> Are the only companies that make sense the ones
... who don't lay off entire arenas full of people and discuss stupid platitudes like replacing them with AI.
Pretty much. Management's messaging around this was so delusional - "replace with AI" - that it's obviously a stock play from a company desperate to cut losses, not a positive statement from a going concern.
If you're this massive inertial machine and the problem is to some degree artificial and a waste of internal resources but someone thinks it isn't, having someone like IBM on the hook gives credibility to the argument that they did their due diligence to solve the problem. Business at scale has a lot of childish fingerpointing. "No one ever got fired for hiring/buying IBM" as the saying goes.
I started my career at IBM in the 1980s. My dad worked there before me.
It's impossible to convey how much hubris existed in the company then, at the peak of the "No-one ever got fired for buying IBM" days. IBM was a god.
The company had never laid off a single employee. The talk was that the R&D divisions had multiple generations of innovative tech up their sleeves and if any competitor ever got close, well they'd just pop something out like from a vending machine to kick their ass. I believed it.
It became obvious when the company started losing the PC market that things weren't actually that rosey at all, and there were serious issues.
Later it became obvious that the company was turning into a consulting company instead of a technology company and the glory days were well over.
Nothing's changed since then. Things have only gone from bad from worse for IBM.
My point - nothing really but disagreement with this observation.
IBM still makes sense, as a dying beast but with customers that can still be milked. As long as those customers can be mikled, IBM will continue to make sense.
It's amazing that this story just keeps popping up over and over. People also said this about Intel (nope, 10nm was a total wash and during that time AMD and ARM started eating their lunch) and Google (no, there is no secret Google lab with the "good AI", it's just Bard and it sucks compared to GPT-4).
The vibe is that they are not releasing stuff until it's trustworthy.
My initial playing with ChatGPT is that it's amazing. Would I trust it to recommend medication? Likely not. Is it useful? Hell yeah.
Google should be taking a good hard look at how they have missed this boat, it could help them flush out their internal arteries.
That's worth something to some people, who really don't give a shit about technology and just want the problem solved.
"We have a great beer customers love? Great, now we're going to use cheaper ingredients. They'll never notice."
"We have great Mexican food, because everything's made fresh on-site? Great, now we're going to shift to a central commissary."
This is great for small companies and startups - if the bigger ones did things properly how could anyone compete?
"if you don't care about something it will be bad" -- maybe to you. To them, it's not causing any problems that they care about.
If one innocently wants to do what's right for the institution, the shareholders or the taxpayers, one has to care. Its usually a tactical mistake to care of course - people who don't have a lot more freedom of action.
Such an ethically questionable decision, trying to push unproven / non-functional tech into life or death decision making.
They were trying to insert Watson into everything. I used to work at a company that worked with building management systems and we had Watson sales people all over our senior managers trying to get 'collaboration' projects going.
Signed, A fan of Robocop as dystopic corporatocracy preview
"Always seize the opportunity to take a divebomb shit on corporate malfeasance through the medium of film quotes." - Wesley Snipes, Passenger 57
https://www.macrotrends.net/stocks/charts/IBM/ibm/number-of-...
What they just announced is a relatively minor round.
Also I am not sure if the 2014 decrease is also related to the formation of GlobalFoundries but I couldn't right away find news about a -80K layoff that year.
They come up with solutions so baroque that it’s exhausting even to hear them talk about it. So you just hire more IGS people to take care of the problems they created. It’s a racket, based on them literally not making sense.
I imagine that getting a board to spend a lot of money on some project is hard and you need the support of the kind of people that can influence board members.
So to get your project past a lot of people who basically don't respect or believe in nerds you had to hire a company with non-nerdy salesmen to back you up.
i.e. the entire thing is social. And just because the nerds (us) have trounced them for the last N years, it doesn't mean that the world has converted into 100% nerd. Shiny shoed salesmen are probably still quite necessary.
I joined as part of an acquisition. I am probably biased but they spent a year making the product multilingual and not fixing anything. Everyone knows the name of the Open Source knock-off of that product but not the name of the product anymore. It's probably still there taking almost no investment of effort.
So I got the impression that IBM is a company where software goes to live out its old age and die. I think of Redhat (perhaps unfairly) with their super complicated Linux distribution which must be a nightmare to maintain because it does so many things and needs to keep long compatibility guarantees. I'm not expecting any innovation whatsoever from them. Perhaps this is very unfair and ignorant of me. I'm happy to be corrected.
Agree with you that many of the salesman are typically technology-agnostic [I use the term "agnostic" in the sense it is used to define religious belief :) ]. They have capable technical guys supporting the salesmen, though.
I think the most horrible bit is selinux - that really upped the level of difficulty a lot. I'm also really fed up with Fedora's frequent releases - so hard to keep your machine current because upgrading usually only works properly once or twice and then you need to do a reinstall. That might be harsh - maybe 3 times but you still end up having to do it or lose updates.
For me this feels liked they've reached a peak of complexity. Now there are flatpacks and other things to torture us with and it's obviously not just them.
I found something that suits me at last and am typing to you from it now. Rolling updates so I NEVER have to do a full upgrade - it's so liberating. I can make choices like not having systemd - I use dinit and it's fine. Development packages are always installed so I want to build something from source and I never have to worry about whether the header files I need are there.
So far I cannot find a negative. OK it was a nuisance to make pipewire start automatically and I probably haven't done it right yet - not good for any new users - but that's it.
your characterization of red hat makes sense from where it sounds like you’re coming from but you miss the mark a bit I feel
RHEL isn’t meant for the everyday developer and red hat does a lousy job enabling them. Ubuntu and Debian do a much better job as is proven by the wide availability of developer tools on Ubuntu/Debian.
But look at the enterprise software market share, of which red hat focuses on entirely — canonical has a sub 5% market share for enterprise Linux or something pathetic along those lines while RHEL is over 85% or something close. Canonical has done the opposite by catering to workstations and developers and have done exquisite things with their workstations but their real enterprise servers suck.
SElinux isn’t meant to just run static on your box while you’re using it to everyday browse or develop that’d be brutal. It’s meant to secure an application or files with predefined egress based on business logic.
Now on the other hand, fedora— it’s upstream my man. Frequent releases is… the point? it’s community based upstream development if you’re looking for the latest features it’s perfect but you’re also getting the latest bugs.
Ubuntu sounds appropriate for you! That doesn’t mean RHEL or fedora “has reached peak complexity”
If you wanted RHEL or Cent/Rocky/Alma:
- disable selinux on your home or personal workstation. you can get granular with how to do it too if you choose. - automate your patching it takes 5 minutes one time - if you want RHEL benefits with Fedora-adjacent (but not as buggy) releases you’re describing what is now CentOS Stream which is also a rolling-release distro
I felt similar to you last year but my company uses RHEL and we just did some worth while quick training. When explained in ways that make sense, it makes sense. Unfortunately I think red hat has been lackluster several times over on how to say what I just said
edit: meant to add that I hope this helps!
I think the whole purpose of RedHat in a way (and of Ubuntu LTS) is the limiting of change.
If you WANT change though, Fedora doesn't compare well to rolling distributions IMO. It has lots of updates but with lots of inconvenience. Rolling distributions just take change continuously forever - there's never a discontinuity so they're very easy to use.
They still mock technical knowledge by wading around in a new sea of acronyms which they don't truly understand or care about. Think of the CEO of a company that makes a build system - every 10th word at one point in time might have been "CD" which means Continuous Delivery which is something his own company never did. But it's the most important thing in the world for that year or two and lots of sententious things are said about it without any connection to the possibility of "walking the walk" to use one of their favorite and least honest terms.
As a technical person one listens to them and wonders what planet they came from and how they could possibly have any connection with the reality of the business they're supposed to be running. But... they're salesmen and have sold themselves as the first product. They know what impresses non-technical people and it isn't the plain honest truth.
I think it boils down to some facts: we have it better than it has been for a long time but there are real situations where one must be a master of bullshit and have the ability to appear supremely confident and all-knowing whilst actually knowing very little. This confidence founded on nothing transmits itself to other people. If you are lucky and appear to be a winner or if you can hide your mistakes and advertise your success then people will take your lead or offer you opportunities.
This included lots of offshoring companies, like IBM, that seemed more about filling out whatever the budget sheet given to them than actually making software or even any evidence of software (commits, issues, etc).
I expect that there are firms out there right now using ChatGPT to produce the facade of work so instead of messing around and billing 8 hours times 30 people every day, they just have a few people fiddling with prompts and checking in the generated text.
I actually think this might have the same impact but would hope that the value just goes back to the client in having employees do the AI work directly.
The joke was that behind every successful offshore project was one or two really busy onshore engineers manically covering and integrating work from the away team.
Relevant: https://twitter.com/ConsultingHumor
We also serve as convenient scape goats in case risky projects go bad.
AbstractBoondoggle boondoggle = new AbstractBoondoggleFactory()As the length of a thread about software design increases, the probability of someone mentioning a "general-purpose tool-building factory factory factory" approaches one.
Including the act of quoting Thor's Law.
———
Though in reality does that mean that the guy bought a metal 3D printer?
To me, ChatGPT's output is indistinguishable from what I've seen offshored coders produce. Only difference is you get the results in 30 seconds in a chat window instead of overnight in a zip file emailed to you.
There are whole businesses here in America that only do rewrites for offshored projects.
After getting ChatGPT to write some code for me, this could happen again except with AI instead of an outsourced team.
So far, it understands the requirements better and writes pretty good code on average. Its unit tests are pretty good too. It always requires some fixes and integration, of course.
Playing into fears about AI replacing large segments of the labour force in the coming years invites calls for regulators to slow down their competition. While at the same time this announcement makes investors happy because IBM is signalling that they're looking to both cut costs and embrace AI.
But, let's say it is. It might play well with investors, but the idea that it's going to spur regulation or legislation to slow down AI development seems optimistic at best. That assumes a government and elected officials who are looking out for workers and not the investors who are going to be giddy about this.
In the U.S., at least, government has been actively trying to suppress wages and maintain higher levels of unemployment. Why would this spur any action to slow down AI?
...and if some employees quit out of frustration about being managed by AI (being "moved between departments" at the whim of an AI sounds pretty spooky to me at least), even better!
Whereas with human managers who knows? Maybe they're good, maybe they're bad, it's a reset every time.
...or your life could become utopia. Who knows?
IBM has long been struggling. They could release a press release saying they’re pausing hiring because businesses are underperforming. Or they could play the news cycle and say it’s because of how awesome their “AI” is. Interpret that as you wish.
Source: https://www.forbes.com/sites/nicholasreimann/2023/05/01/ibm-...
Who need a generative LLM text to print a receipt?
IBM seems to be run by very gullible people.
As are IBM sued for age discrimination, and IBM closes product line
similar to that google guy who left because Google's AI is about to become a sentient, nice little boost to the retirement fund.
They also know it is time to learn to c̶o̶d̶e̶ adapt.
"Alexa...why won't my mainframe IPL?"
IBM is replacing humans, who can identify an exceptional events which weren't anticipated by planners, with AI, which only regurgitates past training data.
I'm sort of imagining IBM human resources department now as some hellish mashup of Ionesco, Nietzsche and Ballard (with a piquant dollop of THX 1138 to provide that absurdist, institutionalized aroma IBM seemed to love.)
At least the TVA was smart enough to recruit biological entities to run it's bureaucracy.
AI tech is being driven by capitalism. The reason companies are so eager to embrace it is to replace people. If you’re in the AI business or advocating for it, you’re (at least in part) an arms supplier for cutting jobs.