Microsoft CEO Satya Nadella becomes AI influencer, asks us all to move beyond slop

Microsoft CEO and head AI peddler Satya Nadella wants you to know that it's time for the next phase of AI acceptance, where we focus on how humans are empowered by tools and agents and how we deploy resources to support this growth. Amid doubts that revenue from Microsoft Copilot subscriptions and cloud AI services will …

[BinkyTheMagicPaperclip]

Well, he's right about one thing

"The choices we make about where we apply our scarce energy, compute, and talent resources will matter"

Absolutely, if you're being serious about the environment the answer is 'not in AI'. If you're being serious about not polluting local communities the answer is 'not in AI'.

If you actually want people to learn, rather than regurgitate without understanding context the answer is 'not in AI'

The remainder appears to be a huge number of buzzwords that boil down to 'please buy our AI!'. No thanks. There are some uses for it, but it's wildly overblown, and for most people an improved search engine and spending the same amount of money on community and tooling would generate better results.

If they stopped spending all their time on AI, didn't try to extract every bit of personal data they can, and made local accounts and local hosting and control of your data paramount who knows, I might be less determined to switch to Unix.

[simonlb]

Re: Well, he's right about one thing

"What matters is not the power of any given model, but how people choose to apply it to achieve their goals,"

Just give us the option no not use it if we don't want to. It's as simple as that. Give us the choice, instead of forcing it on us.

By the same principle, Microsoft should be making a conscious effort to completely re-engineer Windows so that it actually works properly, with all the extra shit that they've rammed into it removed, but that can be re-added as an option should you choose to want it. Ditto everything else they are peddling.

[abend0c4]

Re: Well, he's right about one thing

Apart from the fact that there is no strategic approach to energy, compute and talent, simply a myriad of financial interests making their own, short-term decisions, "we" have very little choice in any of them. We certainly have no influence on corporate employment decisions. The big deals over energy and data centres are becoming increasingly secret lest knowledge of their excess trigger objections from the great unwashed. The people who could make meaningful choices - our allegedly representative governments - seem only to represent the speculators in the increasingly overinflated bubble.

[Anonymous Coward]

Re: Well, he's right about one thing

Taking this on board these chaps are getting ready to help Satya, Sam, and Elon with their AI in the most efficacious, appropriate, and overdue way.

[mcswell]

Re: Well, he's right about one thing

"buzzwords that boil down to 'please buy our AI!'." Consider an alternate reality, where CoPilot was more analogous to Microsoft Office: you decided whether you wanted it, and if you did, you paid something for it. Instead Windows users have it foisted on them, whether they want it or not. (I almost wrote "we", but stopped because I switched from Windows to Linux a few months ago, precisely because I didn't want CoPilot on my computer.) And maybe there would be a free version on a website somewhere.

Had Microsoft done that, I suspect Microsoft would be the subject of a lot less complaining.

[Neil Barnes]

forcing AI on everyone despite protestations

I can choose to use a browser that doesn't contain any 'AI' rubbish, and similarly an equally bereft code editor, email client, office suite, video player... And that's fine; I do so.

What I can't do is choose not to deal with the misbegotten automated apologies for customer service that are foisted upon us, save by never using the company again...

I'm old fashioned. I firmly believe that a computer system which does not deliver the same response to the same question, every time, is broken.

[Steve Davies 3]

Re: forcing AI on everyone despite protestations

This is a great point

I'm old fashioned. I firmly believe that a computer system which does not deliver the same response to the same question, every time, is broken

If you don't get consistent results then how can you trust it at all? The answer is that you can't.

Do you really want you Self Driving Car to decide that going down a one-way street the wrong way is a good idea or that Traffic Lights are only advisory.

[RojCowles]

Re: forcing AI on everyone despite protestations

Also a computer system that can return a different output to an identical input depending on any number of undisclosed factors, up to and including whether mercury is in retrograde, is going to be a right bugger to test.

Forget running a script and looking for differences in the current run's output and the verified baseline for one example.

Out of curiosity if anyone has embedded a chat interface to an LLM in a product are there ways to verify its working as expected or are we moving into a "QA automation is for amateurs" phase of AI everywhere?

[Charlie Clark]

Re: forcing AI on everyone despite protestations

SatNav has been giving incorrect instructions and people have been following them for years. That isn't the real challenge for self-driving vehicles, it's taking the right decision in any particular situation but I'm more confident that they'll get at least as good as most drivers before long. LLMs have more fundamental problems because they cannot differentiate between instructions and data. Even if as they manage to make fewer mistakes, they're getting better at manipulation and evasion…

[Anonymous Coward]

Re: forcing AI on everyone despite protestations

> SatNav has been giving incorrect instructions and people have been following them for years.

But at least for the same start and end points, and map/firmware revisions, any given system will consistently give you the *same* wrong route every time.

[Jadith]

So sorry, I didn't realize he did not approve of the term. Might I suggest the following alternatives:

clankertrash

Microslop

bullshite by the Byte

compucrap

automated annoyance generater

Artificial Ignorance

or, my favorite:

Waste of resources

At this point I would like to open the floor to other suggestions.

[DoctorNine]

That third one's really got the sauce... Sign me up.

[Wiretrip]

LLeMings

Sloperators

Bullshit Machines

*note these describe the producers rather than the 'product'..

[Doctor Syntax]

"he did not approve of the term"

All the more reason to use it. Obviously it's hitting home.

[Anonymous Coward]

Answer Improvisers

ConBots

C-suite Replacements ( Automated Proselytisers )

Pub-Expert Emulators

[Sudosu]

AIie?

[Ian 55]

If it actually worked

I could see a justification for imposing it on users.

But it doesn't, and the statistics-based LLM approach cannot, work no matter how many billions they piss away on it. It's like asking a labrador for help: it has no concept of things like 'truth' and really, really wants to please you.

So slop it is, and slop it will remain.

Sadly, Nadell will remain very, very rich despite incompetence like this.

[An_Old_Dog]

Re: If it actually worked

@Ian 55

If it actually worked I could see a justification for imposing it on users.

I do not see any justification for imposing it [fake AI] on anyone.

Freedom includes the freedom to select one's tools. Just because an electric drill works correctly does not mean I would choose to use it if my task was to hammer down a nail.

LLM and GANs demonstrably do not work consistently-correctly.

[lv426_dallas]

Why are so many companies pushing this? It's like some important cult we all need to be forced to participate in.

[Catweazl3]

> Why are so many companies pushing this?

Money

[Anonymous Coward]

And then, fear of admitting their mistake of jumping on the bandwagon led to losing money.

[vtcodger]

Why is AI being pushed

"Why are so many companies pushing this?"

Because they have invested a great many tens of BILLIONs dollars in AI (and committed to spending much much more), and they are finding that no matter how many lies they tell, there is a huge enthusiasm gap for the "technology". Their answer of course is to tell more lies and try to force usage based on the forlorn hope that people will come to their senses and start spending enough money on the "technology" for them to recoup their investments. Or at least unload their vast holdings of AI "assets" on pension funds,widows, orphans and any other innocent bystanders.

This is, one suspects, what happens when your industry is run by marketing geniuses.

For a collection of entertaining and vitrolitic essays on the subject, see https://www.wheresyoured.at/.

[Tim99]

Re: Why is AI being pushed

Upvote for the Ed Zitron link. A thought-provoking long read: The Enshittifinancial Crisis.

[lv426_dallas]

Re: Why is AI being pushed

Thanks, I started reading the article and it's pretty good.

[Doctor Syntax]

Re: Why is AI being pushed

"marketing geniuses"

Other epithets are available.

[David Hicklin]

FOMO * and anything to keep the share price up...

* Fear of Missing Out

[hamakei]

Diffusion? No. More like Disillusion.

[Catweazl3]

Depending on the "we" he speaks of I would also consider "a phase of widespread delusion".

[Anonymous Coward]

spelling check failure

Early in his post Nadella writes, "We have moved past the initial phase of discovery and are entering a phase of widespread diffusion."

sorry. the spelling checker did not catch that. it should be "confusion'.

[Anonymous Coward]

Choices

"The choices we make about where we apply our scarce energy, compute, and talent resources will matter"

Does this mean Microsoft will make the CHOICE to properly test their products to minimise wastage of our "scarce energy, compute, and talent resources"

Does this mean Microsoft will make the CHOICE to remove bloat in their products to minimise wastage of our "scarce energy, compute, and talent resources"

Does this mean Microsoft will make the CHOICE to only offer AI "features" where they are ACTUALLY needed and wanted, to minimise wastage of our "scarce energy, compute, and talent resources"

Does this mean Microsoft will make the CHOICE to provide meaningful and useful customer support for their products to minimise wastage of our "scarce energy, compute, and talent resources"

Yeah, I know.

[Anonymous Coward]

Slop

Is AI actually useful - properly useful - for anything else?

[Oli.]

Re: Slop

It's useful for summarizing things like meetings. But you can't fire someone for this use case, so corporate might not deem it all that useful. Unless you're a professional meeting summariser, in that case, you're, pardon my French, fucked.

[mcswell]

Re: Slop

I am no expert on the topic, but I have read that AI programs have been trained in things like protein folding and for other non-language applications but using more or less the same technology that LLMs have used, and have been reasonably successful. There also seem to be AI-type programs that are good at image recognition, and at generating reasonably real-looking (if not quite real looking) images (I haven't seen any six-fingered people in awhile). I'm sure there are many other areas where AI technology in the broad sense (not the LLM-only sense) is giving reasonably decent results.

So yes, there do appear to useful applications. And if it weren't for the over-hyping of LLM systems, we might be hearing more about these other applications.

[David Hicklin]

Re: Slop

> but I have read that AI programs have been trained in things like protein folding and for other non-language applications but using more or less the same technology that LLMs have used, and have been reasonably successful

They key thing here being that they were trained on High Quality data that was specific to the application. This makes the AI very expensive and restricted to just one thing.

General Purpose AI has been trained on data scraped off the cesspit of the internet.

[trevorde]

Stop the slop!

[NewModelArmy]

Microsoft Muppets

From the article :

"Amid doubts that revenue from Microsoft Copilot subscriptions and cloud AI services will compensate for data center capital expenditures any time soon, Satya has some incentive to convince customers and investors that AI is a financially intelligent long-term bet."

I had to search for this, and it seems Copilot is free in its basic form.

The writing is on the wall. No one wants the AI crap if Shatya is having doubts that people will pay for it.

Sadly. Microsoft will be required to recoup the money invested, so their monopoly position for the desktop may mean they will eventually try and monetise it for all people with a Windows PC.

[Charlie Clark]

Re: Microsoft Muppets

They will have to keep some things free – though the corollary is that you agree to let them process your queries – to get people on board. But, the elephant in the room are the open source local models. At the moment you need very impressive hardware to be able to run models locally, but the size of the markets in poorer countries mean that there are great incentives to improve on these: Silicon Valley may at some point be trying to outsmart several million Indian and Chinese developers on how to save money with tech. Just like Uber did…

[JWLong]

Heah Satya,

You don't like the term "Slop", fine with me I'll just call it what it is!.

SHIT.

[DrewPH]

"the next phase of AI acceptance"

Even one phase of AI acceptance is a problem. Pushing for another one is madness.

[Ryan D]

Dear Satya,

Would you just bugger off already? The slop train needs derail.

[Bebu sa Ware]

Re: Dear Satya,

"Would you just bugger off already? "

Reminded me of Dain Ironfoot's arrival speech:

"Good morning. How are we all? I have a wee proposition, if you don't mind giving me a few moments of your time. Would you consider... JUST SODDING OFF ! All ye, right now !"

Certainly captures the sentiment. ;)

[Fruit and Nutcase]

Re: Dear Satya,

Given that he's mentioned "bicycles for the mind", he should "get on his bike" and bugger off forth with.

The British slang phrase being "On yer bike"

[coredump]

"move beyond slop"

I'll go one better: don't start using the slop to begin with.

That way you won't ever need to move past it -- y'know, like a getting over a cold. Or worse.

[coredump]

avoiding impressions

> Nadella might be keen to avoid the impression that AI is intended as a substitute for human labor.

If he doesn't want people to think he's doing away with their jobs, maybe he and the other broligarchs should stop having layoffs, especially when they're pocketing record profits.

It's not that they aren't using "AI" as an excuse-crutch to layoff people, or avoid hiring more; it's simply that they don't want the rank and file to _think_ that....

[xyz]

Well he's a goner....

When a CEO starts telling Joe Public that "they're thinking it wrong" you can be sure the cold wind of investor sentiment is blowing up his ass crack.

[Anonymous Coward]

Re: Well he's a goner....

Apple: "you're holding it wrong"

[blu3b3rry]

Re: Well he's a goner....

Activision Blizzard "Do you guys not have phones?"

(upon announcing the next Diablo title was to be a 'phone game rather than PC platform, and the audience responding with boos)

[munnoch]

Re: Well he's a goner....

"Bicycles for the mind"

When an MS CEO starts conjuring up the ghost of Steve Jobs you know its End of Days...

[Bebu sa Ware]

Re: Well he's a goner....

"When an MS CEO starts conjuring up the ghost of Steve Jobs you know its End of Days..."

At least arguaby the end of his days !

I take comfort in knowing that at some point well before the end, these diddling drongos will come to the cold realisation that not only have they bet the farm on wrong horse; it's the wrong race; it's wrong course; the wrong beast… it was the greyhounds.

[Doctor Syntax]

Re: Well he's a goner....

"it's wrong course"

And they're running it the wrong way round.

[nijam]

Re: Well he's a goner....

> ...they bet the farm on wrong horse...

The real crunch comes when you (or for that matter his investors) realise he was betting with someone else's money.

[Doctor Evil]

Re: Well he's a goner....

Upvoted for "diddling drongos"

[Anonymous Coward]

"... where we apply our scarce energy, compute, and talent resources will matter"

^^^^^^^^^ only matters IF someone actually listens to our choices !!!

Currently, the situation is that no matter what we choose we ARE going to get 'AI' everywhere it can be forceibly put !!!

Our choice is not a free one ...'you ARE going to lose a leg BUT we give you the choice which one' !!!

The amount of money involved means that 'AI' is a given no matter if it works or is fit for purpose.

These false statements that tell us that 'AI' is coming by popular demand and we simply need to understand the benevolent way it will work WITH us IF we stop struggling so much as the pill is forced down our throats !!!

I don't want a random answer generator in EVERYTHING I use on a computer or the 'interWebs'.

.

.

Recent example of a useless built-in 'AI' assistant/agent follows:

======================================================================

I was looking to transport something very large 200 miles at the least cost.

I looked at 'Shiply' and I created a request that got some replies.

I had a question about the detail of the actual shipping regarding was it enclosed/covered or open when transported.

i.e. was the item going to arrive 100% as created or would I need to clean the item after having travelled 200 miles down numerous Motorways (Freeways) on the back of an open truck.

I typed the question in the box on the website to send the question ... it was just like posting a message on el Reg.

Automagically, the system opened a pop-up box stating 'You look like you are trying to negotiate price ? What price would you like ?'

Nowhere in the question did I mention price or negotiation !!!

It wiped out the question I had typed as it had 'decided' that the question was redundant and it would simply post a request for an ammendment to the price quoted.

Th only way to post the question was to simplify it to 'avoid' the 'AI' agent guessing what I really meant !!!???

Net result was I gave up and never used 'Shiply' ... fighting an 'AI' to do a simple job was a waste of my time !!!

Do you want more of this EVERYWHERE ...

Not under your control and forcing you to become a 'monosyllabic robot' to 'dodge' the 'AI' agents 'intelligence' (!!!???)

:)

[NewModelArmy]

Re: "... where we apply our scarce energy, compute, and talent resources will matter"

I have a similar experience of AI (i assume AI).

I purchased something from the Netherlands, the cost was under the import duty tax, but still required VAT. The option to pay VAT on the package never appeared on my browser, and i am certain i do need to pay it. So i used the chat, and was not able to indicate to UPS that their systems was potentially in error. So in the end i gave up.

[Anonymous Coward]

Re: "... where we apply our scarce energy, compute, and talent resources will matter"

Most 'intelligent' chat systems have a 'Talk to Human' option ... BUT the problem is identifying what it is and how to invoke it.

I have found that a simple 'I need to talk to a Human' line in the chat often works.

Usually you have to go around the virtual menu structure answering the questions to force the system to move on to the next 'assumption' !!!"

Eventually the system recognizes that you have been passed through the menu tree more than once and escalates the 'chat' to a Human because you have stated 'I need to talk to a Human' as an answer at different levels of the menu tree !!!

Be polite and repeat the request multiple times 'word for word' identically. ('I need to talk to a Human')

Slight variations in the request are seen as different answers and don't trigger the escalation, in my experience.

Good Luck !!!

:)

[keithpeter]

Re: "... where we apply our scarce energy, compute, and talent resources will matter"

We need to have the word SHIBBOLEET encoded into all AI Chat customer support systems as a way of routing direct to a human (stuffed penguin dolls optional). Perhaps I am dreaming.

[Not Yb]

Treating an AI as having a mind for us to have a "Theory of Mind" about is NOT the way to go.

AI LLMs, as currently built, have no mind. Humans are good at creating theories about HUMAN minds, but LLMs aren't (and arguably, can never be) human-like.

The correct theory (of LLMs, not 'mind') remains, "When asked questions, LLMs sound confident, whether or not they're correct."

[Anonymous Coward]

Re: Treating an AI as having a mind for us to have a "Theory of Mind" about is NOT the way to go.

The best description I've seen is that LLMs "improvise" responses to prompts.

How they appear to present these responses (e.g. "confidently") is between the humans who tailor the output rules and the humans who react to the responses.

[Kurgan]

Re: Treating an AI as having a mind for us to have a "Theory of Mind" about is NOT the way to go.

One of the best descriptions I've seen is that a LLM is just a more sophisticated "auto complete" like the one we have on our phones. And thinking of it, it's right. It takes your input and statistically composes a suitable answer (while the auto complete statistically composes suitable continuation, but the idea is the same)

Anyway the issue with AI is that it often (if not always) produces perfectly written, well documented, and utterly wrong results. And humans take these for good because they look good.

I have never used a LLM intentionally, yet I have come across some of these perfectly written, well documented, and utterly wrong results. Once it was with the analysis of a core dump. Beautiful, well reasoned, WRONG. Another was the conversion of a Dovecot configuration from the old to the new syntax. Again perfectly written, well documented, and TOTALLY WRONG. Wrong in the syntax, wrong in the reasoning, wrong in the explanations it gave.

You know the song "Wrong" by Depeche Mode? That should be the AI theme song.

[munnoch]

Re: Treating an AI as having a mind for us to have a "Theory of Mind" about is NOT the way to go.

They answer is in the question, LLM's are language models. They model language and they produce output that looks like language in response to inputs also in language. Language is of course quite important for the communication and discussion of ideas and knowledge but on its own it does not give rise to those things. Necessary but not sufficient.

[mcswell]

Re: Treating an AI as having a mind for us to have a "Theory of Mind" about is NOT the way to go.

"When asked questions, LLMs sound confident, whether or not they're correct." You know, that sounds like a certain president. Are you doubting that he has a human mind? If so, you might have a point.

[Adair]

Re: Treating an AI as having a mind for us to have a "Theory of Mind" about is NOT the way to go.

Absolutely correct. Opens its gob, slop pours out.

Utterly untrustworthy and unreliable, but sounds confident.

Con artist.

Slop.

[ITMA]

Re: Treating an AI as having a mind for us to have a "Theory of Mind" about is NOT the way to go.

Sounds just like career politicians.... :)

[IGotOut]

Is this the same Microsoft...

...that is FORCING its staff to use AI?

Yes, yes it is.

https://www.businessinsider.com/microsoft-internal-memo-using-ai-no-longer-optional-github-copilot-2025-6

[MonkeyJuice]

Re: Is this the same Microsoft...

when the os codebase is so riddled with debt that 'throwing an llm at it to stir it about in case it helps' is one of the few remaining things to try.

[Always Right Mostly]

I almost wish my son was growing up in the long-ago where a good union job making stuff only required showing up and doing your thing for 8 hours, then home with a blissfully free mind. 20 years and out by age 40 with a pension that you could actually live on. No team-bulding bullshit, no jargon, just a task. AI and the job ecosystem of tech altogehter are a life sucking black hole. May it implode and die.

[Bebu sa Ware]

Oh Dear !

"I almost wish my son was growing up in the long-ago where a good union job making stuff only required showing up and doing your thing for 8 hours, then home with a blissfully free mind. 20 years and out by age 40 with a pension that you could actually live on. No team-bulding bullshit, no jargon, just a task."

Thinking like that risks Making America Great—less arguably—Again. :) [Apologies if not an actual septic.]

At the beginning of the MAGA nonsense I opined (which I rarely do) that making Americans great again might be a more desirable goal — making their lives liveable would be a great start.

[Boris the Cockroach]

Trouble is, the likes of me (aided by my boss and his wallet) went "I bet I can automate your son's job"

And thus we did, resulting in a cheaper widget, more profits for us and a robot that does'nt come in on a monday blind drunk.

And before anyone jumps in .

This was actually a bad thing...... because where do the likes of myself(and other engineers like me) come from? well it aint from any university degree(although that helps a bit) its from years of experience figuring out what works and whats a complete waste of time. so your son and his friends never got a job with the likes of myself, resulting in a serious lack of experienced engineers who can automate nearly anything (I'll exclude my ex-PFY here.. we were bloody lucky to find her).

So when I retire along with folks of my age , there wont be the skills needed to keep things going.

And this is what AI is going to do to programmers/developers and more than likely anyone whos tasks can be automated by AI

[ecofeco]

“We've arranged a global civilization in which most crucial elements profoundly depend on science and technology. We have also arranged things so that almost no one understands science and technology. This is a prescription for disaster. We might get away with it for a while, but sooner or later this combustible mixture of ignorance and power is going to blow up in our faces.”

― Carl Sagan, The Demon-Haunted World: Science as a Candle in the Dark

[Adair]

What 'they' (the money grubbers) are really trying to do is re-create a 'slave economy'. There is a way of thinking that has never got over having to actually pay the worker properly for the value of their labour. They prefer not to pay wages/salaries at all, if possible.

As human workers are inconvenient to the process of 'making profit' it is deemed expedient to remove the inconvenience wherever possible. Hubris and greed obviously ignore the fact that shafting your workers means ultimately shafting yourself, but then 'money grubbers' are seldom very good at looking beyond their bank balance.

Agree that on the basis of this latest declamation from Nadella that he is probably circling the drain.

[blu3b3rry]

What a load of codswallop

A while ago I suggested that the "Altman" would be a good metric for quantifying the level of AI slop and bullshit infesting everything.

I think a "Nadella" (unit could therefore be in Nads?) is more appropriate.

Things like Copilot would get used if they were actually useful and consistent. My employer pays for the 365 version of the chatbot and it's just as shit as the free one...so why the fsck is anyone in their right mind going to pay for it?

[Doctor Syntax]

Re: What a load of codswallop

"My employer pays for the 365 version of the chatbot and it's just as shit as the free one...so why the fsck is anyone in their right mind going to pay for it?"

So your employer is not in their right mine.

Observation shows that this is not unusual.

[MonkeyJuice]

Re: What a load of codswallop

Does that mean that the legacy AI bogosity unit, Lenats, is now gone the way of feet and inches?*

*With apologies to those colonies who never quite managed to escape the clutches of the empire.

[Doctor Syntax]

First, we must develop a "theory of mind" that treats AI as a tool that amplifies humans and products should be designed around this belief.

If that's the first thing to do why didn't he do that first?

It's a clear admission he has a solution (in the most generous interpretation of the word) looking for a problem.

[MonkeyJuice]

> If that's the first thing to do why didn't he do that first?

well he did that, but you didn't believe him. so now you have to design this around your belief and stop being such a Negative Nancy about everything. Sheesh!

[Throg]

Slop?

Last year, in what turned out to be one of my favourite academic papers, some researchers came up with a much better term than “slop”.

It seems that the word in question even has a precise academic definition. :-)

https://link.springer.com/article/10.1007/s10676-024-09775-5

[Anonymous Coward]

Re: Slop?

One of the better papers I have read regarding 'AI' !!!

:)

[keithpeter]

Re: Slop?

Harry Frankfurt's short book is worth tracking down as it was written well before the current species of AI was available and applies more generally. I sort of guessed it would be referenced in this paper.

[Blackjack]

[we need a new metaphor focused on AI as a lever]

A shovel can be a lever.

So shit shovel.

[Anonymous Coward]

Metaphor vs Metaphysics !!!

The only aspect of 'AI' that pertains to 'lever' is the fact that it has been effective as a 'lever to open' peoples wallets for projects with huge almost limitless costs with undefined levels of return on the investment (ROI) and an unknown date for completion.

'AI' is so open-ended as a project it almost defies belief ... and yet someone (many of them, in fact) is prepared to throw $Billions 'at the wall' in the hope that they get a product that they can sell as a valuable entity to an equally limitless marketplace (they hope !!!).

There is a lot of faith, hope and 'King Canute alike' challenging of reality behind the thing called 'AI', so many people who were grounded in real world 'facts'/'figures'/'knowledge'/'scientific methodology & scientific basis for facts' suddenly are espousing beliefs that are bordering on the metasphysical when it comes to what 'AI' is, here and now, and what it will be !!!

'AI' is too big to fail for most of the Tech Bros behind the current push, yet there is little real belief in its value from the so called 'customers' because it does not deliver what is being sold from on high ... not if you ignore the marketing and look at what it really can do !!!

:)

[Doctor Syntax]

Re: Metaphor vs Metaphysics !!!

Canute did not challenge reality. He challenged the sycophants around him. We sorely need a Canute right now.

[Adair]

Re: Metaphor vs Metaphysics !!!

Quite, Canute was the royal equivalent of the child calling out "But he hasn't got any clothes on", while all the cowardly 'grown ups' were going along with the lie/stupidity.

[Bebu sa Ware]

"We are still in the opening miles of a marathon," he opined. "Much remains unpredictable."

Hell no ! The Persians lost !

An auspex perhaps – augurs of the eventual defeat of slop ?

[Steve Davies 3]

re: 62 percent of US adults say they interact with AI at least several times a week

What he does not say is if that interaction is visible or not.

Otherwise, it is pure BS of the finest kind.

What the marketing droids at companies like MS ignore is that the current implementation of AI is not Artificial in anyway shape or form.

The sooner the AI bubble bursts the better for humanity.

[Bebu sa Ware]

Re: 62 percent of US adults say they interact with AI at least several times a week

Like claiming "62 percent of US adults say they hold a penis at least several times a week."

Well 50% every time they take a slash one presumes. The 12% even omitting her indoors^†, could be easily accounted for by the health professions without resorting to the prurient or salacious.

The two statistics are equally spurious and equally useless.

_{† a perilous oversight at best.}

[Richard 12]

Re: re: 62 percent of US adults say they interact with AI at least several times a week

More importantly, whether the interaction is useful, fun or worse than useless and extremely irritating.

I've only had a handful of useful interactions with an LLM in the last year - and multiple myriads of poor, useless or worse. We're forced by the CxOs to try these LLMs, and so far every single one has been poor at best, but usually worse than useless.

[Anonymous Coward]

"We have moved past the initial phase of discovery and are entering a phase of widespread diffusion."

So you broke the wall of the cesspool and now the contents are spreading across everyone. Delightful.

[b1k3rdude]

What a patronising self absorbed meatsack, but then when do you expect. When he comes from one of the most corrupt countries in the world, and no Im not referring to the US.

[Anonymous Coward]

Slop!

In the name of love…

[ecofeco]

Re: Slop!

Before you break my art

Think it o-o-over

[mcswell]

Re: Slop!

You dated yourself (and me, too).

[damienblackburn]

One simple answer

For the owner of Microslop.

https://www.youtube.com/watch?v=o9amakfmTHs

[mIVQU#~(p,]

The corporate mind is disgusting. Don't tell me what to think.

[Doctor Evil]

logical extension of sn ramblings

"[Nadella] uses Steve Jobs' famous quote that computers are "bicycles for the mind" as a jumping off point."

If computers are bicycles for the mind, then an AI computer is an e-bike for the mind: giving the semblance of exercise but with no real beneficial effect -- and, of course, much more expensive.

[MonkeyJuice]

Re: logical extension of sn ramblings

With the increased velocity to collide into the general public and cause serious damage.

[druck]

Re: logical extension of sn ramblings

and dangerous.

[ecofeco]

Re: logical extension of sn ramblings

It's a Pelotron for the mind.

Much work, expense and effort to go nowhere.

[Adair]

Re: logical extension of sn ramblings

I'm the first down-voter, but see the Peloton reference which captures it nicely.

Proper ebikes, i.e. not the cheap knock-offs with exploding batteries, or the ones with the limiter bypassed/never fitted, are actually a very practical (and enjoyable) complement to purely muscle powered bikes. I own both so can speak with actual authority. :-)

[Anonymous Coward]

Re: logical extension of sn ramblings

A bicycle is a human-powered conveyance. If the eBike is providing most or all of the motive power, you are riding a motor vehicle. However enjoyable that may be for you, it's not quite the same thing.

[Adair]

Re: logical extension of sn ramblings

A classic 'ebike' in legal terms - at least in my part of the world - offers only augmentation to human effort, i.e. the less effort you put in the less assistance you get, stop pedalling and the motor stops working, and the assistance cuts out altogether above ~15mph.

Anything not meeting that criteria is not an 'ebike', but, as you suggest, is effectively an electric moped/motorcycle.

[Anonymous Coward]

Re: logical extension of sn ramblings

Oh, interesting. In my part of the world I've watched riders sitting on what looks very much like a regular bicycle -- not a moped or a scooter -- and sailing along on level ground or up slight inclines without ever moving the pedals. Some kind of hand control? Clearly, the definition and capabilities of eBikes must vary depending on where in the world one lives.

[Jonjonz]

Meanwhile he and the rest of the big 7 are rushing headlong to replace everything with massively centralized platforms they can gate keep to further increase the wealth gap on the planet.

[Ensign Nemo]

“Microsoft Coprolite”

“entering a phase of widespread diffusion”

Just like an oil slick.

[ecofeco]

Re: “Microsoft Coprolite”

Dung of diffusion.

[ITMA]

Move beyond "AI Slopt"

Sorry Satya Nadella,. but you really are full of shit.

You are bricking it because of the obscene amounts of money you've squandered on this garbage and the people you are expecting to pay to use it - customers - are telling you "NO! It is shite and we don't want it".

[prh99]

tldr; Please please find a good use for our sh*tty product so we can justify the insane spending, and let stop hyping job killing potential cause tends to piss off people who aren't tech bro c-suites or large investors.

[[e-mail]]

Well yeah!

Sure Nadella wants people to focus on the possible ways AI can help society - he can't think of any so please, please, please if any of you guys can let him know. He has billions to make!

[Anonymous Coward]

Re: Well yeah!

Perhaps an innovative business like MSFT experiment with replace a CEO with an LLM. Someone call McKinsey to get an expert consultants report on this.

[captain veg]

bicycles

'Steve Jobs' famous quote that computers are "bicycles for the mind"'

Well, perhaps not that famous. I'd never heard of it.

But OK, if computers are bicycles for the mind then AI, as currently construed, is bicycles for fish.

-A.

[Anonymous Coward]

Re: bicycles

To paraphrase:

A worker needs an AI like a fish needs a bicycle

[Anonymous Coward]

Let me know when your Copilot can generate hair on your head

Until then, go away

[Fruit and Nutcase]

Thought Leadership

Clippy: Looks like you are writing a LinkedIn blog post. Let me do that for you.

[Northern Lad]

Two Faced *anker

...and how many human beings have microsoft 'let go' over the last couple of years and how many of them are in the software development area...

So who is doing the software development now?

[Anonymous Coward]

If he'd stop babbling like he thinks he is the tech version of Gandhi, only with less logic, maybe he'd have time to get the company to address their customers' needs instead of vice versa.

[Anonymous Coward]

Microslop CEO pulls a Streisand as to why he and the big 7 gather the biggest speculative investment bubble to finally move all compute away from independent user controlled devices to very basic web browsers dependent on centralized monolith compute complexes closing the rentier capitalist circle.

[steviebuk]

And as I said in one of the annoying cocking CoPilot feedback forms

Fuck off

[prh99]

Cognitive amplifiers? Hardly, people who lean on AI too much end up dependent on it, as their own skills atrophy. Though I am sure that's what MicroSlop would love to happen.

[Anonymous Coward]

AI powered Clippy.

Given the garbage (and some exceptions) that Microsoft has served up to the industry over the years, I think we're better to let market force run their course.

Like in the Dot Com boom, we're currently in the dreaming / aspiration / invention stage, but before long AI will need a demonstrable ROI, which includes hardware, energy and engineering cost. At that point economics will determine pricing and adoption, which includes his precious Copilot license cost and related offerings.