Recently, senior executives at Salesforce have admitted, both internally and publicly, that they massively overestimated AI’s capabilities. They have found that AI simply can’t cope with the complex nature of customer service and totally fails at nuanced issues, escalations, and long-tail customer problems. They even say that it has caused a marked decline in service quality and far more complaints.
But the problems go far deeper than that.
Both employees and executives have said that the company is wasting countless resources on firefighting to stabilise operations since the mass AI layoff. Employees have to spend so much time stepping in to correct the wildly wrong AI-generated responses that AI is wasting more time than it saves. In other words, this AI reduces productivity, not increases it.
But there is also a huge problem here with expertise and skill debt. On top of the firefighting to correct the AI, executives have also highlighted how they are also having to firefight to stabilise their systems from problems that were previously easily solved by staff who had the required experience and skill. However, these staff were fired in the AI layoffs.
Expertise, experience and skilled employees are really hard for a company to acquire. You see, much of the expertise, experience, and skills required are unique to the company and its operations. These operations will have quirks, common problems, and unique issues that even the most experienced outsider will really struggle with, but are effortless to someone with experience within the company. As such, these attributes are not only vital, but are nurtured and grown within a company, and cannot be hired in on a whim. What Salesforce has done is chuck all this experience out the window, and now they are suffering.
To say this is a damning indictment on Benioff, his capabilities as a CEO, and his cult-like push for AI is an understatement. It demonstrates that he is dangerously ignorant, as it was painfully evident that this would happen; more on that in a minute.
As some of you know, I build software for a living. AI has been a productivity godsend for me, but I use it in a very targeted manner. For example, the past few days, I’ve been writing automated tests for a large piece of software I’m building. With some back & forth, an AI can build a suite of tests in 10 to 15 minutes, what used to take me 4 hours to a few days to write manually. But I’ve learned to take a stance of “don’t trust and deeply verify” because AI is simply nowhere near as qualfiied as a human expert on the subject at hand.
Industry leaders, such as the folks at Salesforce, have been sold a bill of goods. The old adage of “B.S. Makes the World Go Round” has never been more true than when it comes to AI.
Anything generated automatically must be verified. For example, an integrated circuit goes through several levels of abstraction, or implementaions or formats, from high level design to physical netlist, each level being generated automatically from the previous. It has to be verified at each stage, as do the verification tests themselves. At some point, the human has to give the OK.
Industry leaders, such as the folks at Salesforce, have been sold a bill of goods.
I could have saved them millions, if not billions, of $$$. Idiots…
What’s that old saw? There’s a sucker born every minute? Nice to see the 21st century continues the fine tradition.
I work with AI as well. AI is an augmentation tool, not a replacement. The way a table saw made a carpenter or cabinet maker more productive without replacing the person using the tool.
I have a friend who does software for … something. He says that he is using a lot of Ai now, but it’s all about saving time rather doing the job. He says that there are jobs that used to take days that he gets the Ai to do and then just goes over for corrections.
He also says that current Ai could not replace him.
As I say, AI IS BS
This happened at the company I do most of my contract work for. My contact at the company was really hyping AI and wanted it to take over large parts of the job we do. To be fair, if it worked it would have been great.
He built an app to test it out and I told him it simply would not work. I would have had to spend so much time correcting it and adding things it missed that I might as well just do it myself. I haven’t heard much about AI from him since then.
The worst of AI.
Try B.C. hydro.
They force you to listen through their BS about how great DEI is and if you become abusive they will cut you off.
The DEI hire through AI then proceeds to abuse you so that you have to hang up before you get to the reason for your call.
I guess through DEI AI it works. It must cut down the number of complaints it receives.
Obviously what Salesforce needs is more AI.
AI me harder daddy!
Very simply put, the customers must be wrong. There isn’t anything wrong with the AI. Can’t be wrong.
Reminds me of either BritRail or the London busses years ago. They did an examination as to why they couldn’t keep to schedule. The answer? Riders were they problem?!?
Ran into that problem with a skip the dishes order… it was completely wrong but even after sending them pictures, they claimed it was the correct food… took more than an hour to get them to refund it and 2 agents after I asked the first one if they were AI or idiot.
It’s astounding that Salesforce apparently didn’t “beta test” their AI customer service before scaling it.
Cutting staff, cutting costs is soooooooo beloved by Wall Street … that the stock price instantaneously spikes upward … triggering bonuses for all in the Executive suite.
Short term, short sighted, planning is no way to operate a business.
Benioff is a big San Fran lib. So I’m rooting for his failure here.
I swing a hammer and can’t comment on the quality of the AI.
But the investment amounts in AI are astronomical.
There is nothing historically comparable.
I don’t know that this guarantees a sudden breakthrough or just a long ascension, but like the airplane or the personal computer or the telephone, its going to get better.
A lot better.
“…its going to get better.”
Or it isn’t, and all that money is wasted…
I’m betting it isn’t. The current crop of AI cannot -reason-. As in, it cannot follow a logic chain. It can’t do mathematics. It SIMULATES following a logic chain, but it can’t actually do it. It kind of wanders off into the weeds after a while.
It can build things in software, but only what you’ve told it to do, AND you have to stand over it while it does it. Otherwise that thing you asked for won’t work.
If they manage to make something that can do basic arithmetic on it’s own, without having it’s little hand held, then I’ll believe maybe all that money wasn’t wasted.
By the way, all the money is going for server farms and cloud storage. Server farms and cloud storage can be used for other things than AI. For example, a no-sparrow-shall-fall social credit system. Like what they have in China.
“John Spartan you are fined one credit for a violation of the verbal morality code.” Demolition Man.
Who would have thought that the two movies best representing this timeline would be Demolition Man and Hunger Games? Seriously.
WIRED magazine used to have an annual “which movie best predicted the future” review and they eventually quit because it just kept being Demolition Man every year.
If true, that’s hilarious!
Budddy / Phantom …. last week I was watching BNN / Bloomberg and the analyst said that AI had been vastly over sold & over promoted. That is typical of anything out of Silly-con Valley IMO but even more so with AI.
The analyst was speaking to the the investing future of AI … ie like so many other high tech promotes, AI was likely to go sideways or even have a minor correction as the hype subsides.
Keep ur stick on the ice folks.
I’m hip to your jives…
But I don’t necessarily mean the Large Language Models that even seem illogical and stupid to me, but maybe the World Model like Yann LeCun talks about? or something(s) else altogether?
“In other words, this AI reduces productivity, not increases it.”
Are you saying that a brand-new technology failed to live up to it’s hype? How can this be?!!! C’est impossible!!
Hate to break it to y’all but AI is here to stay. And yes, it’s had some setbacks through the years but it’s getting better by the month. So it’s a dismal failure with respect to “corporate customer service” BFD. Everyone needs to zoom out. Any application that can sift through Everest sized mountains of data in minutes if not seconds has business owners and analysts licking their chops. And that’s just one of a multitude of skills.
I’m not saying AI will be bricklaying anytime soon. However…
Can’t stop what’s coming.
“…this ignorant AI overconfidence and dramatic deployment, followed by a rapid walk back, is happening across the entire economy.”
Don’t be such a drama queen. Not everyone is diving into the AI deep end.
And speaking of technological advancement – I’m old enough to remember being taught computer punch cards and the slide rule in highschool…fast forward 40 years I’m using an Iphone to find the best restaurants in Italy and buying Florence to Sienna train tickets.
Apple Maps has gone from steering me down the wrong street to giving me advance instruction as to which of the 5 freeway lanes to make my next turnoff.
However! I struggle with my wife’s TOTAL reliance on the verbal instructions… never LOOKING at the map, so she can visualize the pathway ahead. I ALWAYS look at the map before taking the verbal directions.
There’s a clue in that as to how AI can be a helper … but not make us a blind dependent.
Totally with you. However – I don’t know about anyone’s else’s experience in Italy but the inhabitants there remain absolutely glued to their culture/heritage, in that the English language is for other Eurochumps to learn. Not for them.
Their attitude: We love to have you come and visit us to experience everything Italy has to offer…but you best have your A game if you need to ask for directions or anything else for that matter. And I do mean anything. I honestly thought they were putting me on at times…maybe they were.
At the very least, one requires a rudimentary knowledge of the language, sadly I didn’t have it and I can’t say I blame them…it’s their show. The whole world doesn’t revolve around us Anglos.
That said – buying train tickets with the Iphone as opposed to explaining to the Italian ticket guy for ten minutes as to what I want is a godsend.
No fuss no mess.
Is it too late for me to visit Italy for the first time at age 70? You make it sound as though Italy hasn’t yet been fully Islamocized like the rest of the EU. But I’ll need to sharpen my Italian past “Basta” and “Al dente”
What’s the Italian word for; “fish head pasta”? Asking for a friend with fish head allergy.
The reason AI is being shoved down our throats prematurely
is because Liberals have never been able to “cope” with customer service.
Revealing their perspective of the customer,
who’s merely a wallet to be pillaged and quickly pushed out of the way for the next victim.
Management always knows best, and experienced line staff and long-term employees are expendable, regardless of the outcome for customers.
Management not appreciating how much skill and experience their employees bring is a problem as old as the hills.
And just as alive with the sound of music.
Just like the internet, AI is going to change everything.
But just like the internet in 2000, companies and investors should not bet the ranch on it too early. Take a cautious, skeptical approach. It’s going to have weaknesses that you don’t expect.
I’m also old enough to remember when shitty software developer Bill Gates… before becoming a world renowned Physician/Scientist/ Immunologist, proclaimed to the world 640K was all anyone would ever need.
Although whether or not he ever said such a thing is debatable.
He did, but the context has been lost because programmers are bitter, jealous children. That’s why they all hate Microsoft and Bill Gates, despite companies like Sun and Oracle being far, far worse and the most egregious behaviour by Microsoft happening long after Gates left the organization. But programmers do need their Emmanuel Goldstein.
What Gates said was that due to the limitations of the hardware at the time, 640KB of addressable memory was the most they could provide in DOS. And that some day those hardware limitations would certainly be removed, but until then 640K ought to be enough for anybody. But with the full context it’s harder to feel smug and superior about someone vastly more successful.
Hating on Gates and the reasons why deserve their own thread. Or a novel. I’m old enough to remember Microsoft, under Gates, using their monopoly to force competitors out of business. Or buying competitors and shutting them down. Or simply ripping off their products and strangling them to death by delaying court proceedings. At one point Microsoft Microsoft was an excellent example of how badly a monopoly can operation. Think Canadian health care but for software.
A.I?
Definite your terms.
Arrogant Incompetence?
Oh yeah.
As long as we have run away bureaucracy,arrogant incompetence is here to stay.
We have already seen the wisdom of our intellectual superiors,those who have never grown a crop,built a physical thing or managed a project for real..
We have safetyism galore where absolute idiots inform the tradesman how we shall perform our chosen trade.
Yet you cannot tell these experts anything.
Abandon them to their “surprising results”.
You know ,those results we told them would logically occur from their demands.
Artificial Intelligence?
Show me some natural intelligence..
What defines this Artificial Intelligence?
A Turing test run by minions?
For a computer program can convince a utter fool every time..does not mean it’s output is intelligent.
Seems we have one A.I defining the other A.I.
One of the social risks of specialization,is inferred competence outside of ones specialization,which university experts constantly demonstrate.
Do we have real machine intelligence,or is this pattern recognition?
Which can be easily fooled by inputting patterns of fakery..
“The computer says” is our latest version of “God says” for the gullibles.
Salesforce is just the first of many,embracing a fad in the name of increasing stockholder profits,alienating their key staff and wondering..”What was wrong with those employees? Why do they hate us?”
It’s also been suggested that a lot of companies are laying off staff because of financial issues and using Ai as an excuse.
Fire people over Ai and your investors don’t panic.
We don’t need no self-driving pizza.
https://www.foxbusiness.com/lifestyle/video-captures-train-tearing-apart-food-delivery-robot-stuck-miami-tracks
AI is like socialism. The version that works just hasn’t been tried yet.