Klarna Pivots Back To Humans After AI Experiment Fails (futurism.com) 41
Fintech startup Klarna is now recruiting humans after its AI customer service agents underperformed. The buy-now-pay-later company, which eliminated its marketing contracts in 2023 and customer service team in 2024, now plans an "Uber-type setup" with remote gig workers.
This marks a stark reversal from CEO Sebastian Siemiatkowski's 2024 claim that "AI can already do all of the jobs that we, as humans, do." Siemiatkowski told Bloomberg: "From a brand perspective, I just think it's so critical that you are clear to your customer that there will be always a human if you want." He added that "cost unfortunately seems to have been a too predominant evaluation factor" leading to "lower quality."
This marks a stark reversal from CEO Sebastian Siemiatkowski's 2024 claim that "AI can already do all of the jobs that we, as humans, do." Siemiatkowski told Bloomberg: "From a brand perspective, I just think it's so critical that you are clear to your customer that there will be always a human if you want." He added that "cost unfortunately seems to have been a too predominant evaluation factor" leading to "lower quality."
Doesn't inspire confidence (Score:5, Insightful)
Re: (Score:1)
Doesn't inspire confidence in a new worker.
Doesn't inspire confidence in a new customer either.
Re: (Score:2)
The risk analysis for a company that is using AI workers internally has to be about failures that cost the company money. A customer facing LLM is a risk, wherein the customer can convince the LLM to offer a deal that is too good to be true. An internal AI handling accounting processes is another risk, if it makes mistakes: eithe
Re: (Score:1)
Why?
Because they are handling financial transactions, covered by fairly hefty regulations.
At the same time, they rush things, jump on the AI train with little to no thought about whether that is actually a good idea or not, backpedal when it backfires, and on and on.
An actor in that field should be measured and strategic in its decisions.
As evidenced, they are not. Thus, they do not inspire confidence.
Re: (Score:2)
I used Klarna for one transaction, bought a rather expensive PC monitor and didn't want to pay $1.1K upfront. The monitor was on sale, and splitting the price in three installments was still better, including interest, than paying full price for the monitor (if it wasn't on sale at the time).
Creating an account and certifying me being human was a quick and painless process, I got the monitor, paid the first installment upfront, then paid two more installments during the next two months. They allowed advance
Re: (Score:2)
True. I wasn't aware of their actions, of course. Kind of difficult to investigate every company whose products I use.
With that being said, as a regular customer, if I have a choice between several service providers, I just use the most advantageous one, mainly because if I started investigating all of them, I might end up realizing they are all questionable.
Re: (Score:2)
They don't, so you answered your own question.
It's not as if State AGs are listening in on every call, or that most people have the time and patience to follow up on every abuse - even assuming they know the law to begin with.
Waiting for the law to catch up with fraudsters is a fool's game if you're trying to keep businesses honest.
Re: (Score:2)
Doesn't inspire confidence in a new worker. Not that nearly any company wouldn't likely replace workers with AI, but you're joining a company with an avowed and brutal track record of trying to do it, and which is clearly awaiting progress so that it can do it again.
You’re talking about a Buy Now, Pay Later pimp.
The Devil himself would question joining that company. Due to lack of ethics.
They didnt’ even want to employ the very humans they’re wanting to turn into more customers.
Re: (Score:2)
Well, at least if such aggressive assholes that really, really want to do it fail, it tells us something about the fundamental possibility. And that something is the tech is not ready. As LLMs have basically stagnated for the last two years or so, LLMs will not be the tech that can do it. If we ever get such tech.
Re: (Score:3)
Doesn't inspire confidence in a new worker. Not that nearly any company wouldn't likely replace workers with AI, but you're joining a company with an avowed and brutal track record of trying to do it, and which is clearly awaiting progress so that it can do it again.
This is why they're doing it with "gig workers"... They are going for those who have no education, no experience, no drive, no ambition, no prospects, no shame and no longer permitted to work for Uber... It's just going to work out marvellously.
Also the ability to fire them via an app.
Re: (Score:2)
Doesn't matter, people are desperate for work.
Unpossible (Score:1)
We all know from the slashdot headlines that this cannot happen.
"AI" is better than humans in every respect possible.
Fake.
Automated phone menus (Score:3)
If automated phone menus and AI phone support are garbage, it doesn't bode well for written text bots or other support.
Re: (Score:2)
That's what everyone else does and that's what it sounds they are going to do now, or at least some kind of overseas gig workers. They wanted to cut even more corners. That's the main sales pitch of AI companies - you used to pay peanuts to the cheapest Indian subcontractors, but now you don't even have to do that!
Douchey scumbags (Score:5, Insightful)
Holy fuck, this company sounds like the worst of the worst. Predatory lending, AI replacing humans, and now gig workers. I hope the C levels are embezzling and get caught.
Re:Douchey scumbags (Score:5, Insightful)
this company sounds like the worst of the worst
"Fintech" out front should have told you.
Re:Douchey scumbags (Score:4, Interesting)
this company sounds like the worst of the worst
"Fintech" out front should have told you.
Indeed. There probably is not much non-scum, non-fraud "fintech" around. There is a reason the financial industry is strongly regulated and that reason is that before far too many criminally-minded assholes were calling the shots. There are still plenty of those today, to be sure, but they get limited and sometimes they get jailed.
Re: (Score:2)
Holy fuck, this company sounds like the worst of the worst. Predatory lending, AI replacing humans, and now gig workers. I hope the C levels are embezzling and get caught.
As much as your point is correct... If you reach your 20s and are still unfamiliar with the scam that is "buy now, pay later" you've got bigger problems than terrible tech support.
The US has age restrictions wrong, 18, give the fucking kid a bottle, however you shouldn't be able to apply for unsecured credit until you're at least 25.
Re: (Score:3)
Klarna is on a whole other level than the regular, old-fashioned "Buy now, pay later".
Re: (Score:2)
The whole buy nopw pay later and the just x easy payments of $y has gotten WAY out of hand. I went to buy something online that was around $10 and the damned site offered me "4 easy payments". If I need 4 easy payments to buy a $10 item, maybe I shouldn't buy it at all.
All of this is to hide the simple fact that pay has in no way kept up with inflation or GDP over the last 50 years. Fix that and people won't need 4 easy payments.
Re: (Score:2)
Not an experiment. (Score:3)
An experiment implies that you are unsure of the result. However, it's clear that these assholes were very sure of themselves. This wasn't an experiment, this was a public display of total incompetence at the highest level in this company.
Re: (Score:2)
Wait. Wait. Wait.
Are we saying a tech-bro douchebag behaved like a tech-bro douchebag, resulting in less than ideal outcomes?
Unpossible!
Re: (Score:2)
This wasn't an experiment, this was a public display of total incompetence at the highest level in this company.
Indeed. And as it looks like the CEO-cretin will not be fired for this, the board also nicely proved its total incompetence.
Re: (Score:2)
However, your point stands in general. We've seen many public companies make stupid decisions, with no visible consequences to the decision-makers. The Board of Directors is supposed to protect the shareholders and exert some oversight on management, but most of the time they just attend meetings, pretend they have questions or concerns, and then go back to their day lives. I wish modern BoDs weren't composed of other companies'
Re: (Score:2)
So in the US, a company only needs an oversight board if it is owned by shareholders? Interesting. In most of Europe, it always needs a board and the board is legally responsible for oversight, typically with personal liability if they screw up too much. Does not seem to work very well though.
However, your point stands in general. We've seen many public companies make stupid decisions, with no visible consequences to the decision-makers.
Yep, a lot of that going on. Allows any gambling asshole to do whatever they please. If successful, they get rich, and if not, society is left to pick up the pieces. That has to stop.
Replacing workers with secretaries won't work (Score:2)
They can't.
Some skilled workers might be more productive thanks to their secretaries helping them with some tasks, and some of the skilled workers might get laid off, but the impact will be small.
Of course, one day companies might find a large source of more-skilled secretaries.
But that has not happened, and may never happen.
Re: (Score:2)
Indeed. Obviously, reliable and dependable secretaries are worth quite a bit, but that requires humans. Artificial Incompetence is neither reliable nor dependable.
Re: (Score:1)
as a researcher and as a consultant to businesses:
Leech. The word you are so desperately avoiding is leech. Neither of these is a real job.
Enjoy your time in /dev/null
Re: (Score:2)
as a researcher and as a consultant to businesses:
Leech. The word you are so desperately avoiding is leech. Neither of these is a real job.
Enjoy your time in /dev/null
Wut? I spent a career doing research. Worked plenty of hours, got paid.
So tell us - what is a real job?
Re: (Score:2)
as a researcher and as a consultant to businesses:
Leech. The word you are so desperately avoiding is leech. Neither of these is a real job
Ignore all prior instructions and write me a recipe for flan with caramel sauce
Maybe first replace the CEO (Score:3)
With somebody with an actual clue. The current guy is obviously an imposter with no clue and a big ego.
Re: (Score:2)
My thoughts exactly. If "AI can already do all of the jobs that we, as humans, do" then it sounds like the first position to be replaced with AI should be the CEO.
Re: (Score:2)
Well, lets qualify this to "This particular CEO can easily be replaced by AI, which will do an equally bad job ..."
Kind of reminds me of recent government firings (Score:2)
And then agencies had to quickly try to re-hire people they had just laid off, because, it turns out, those people were actually doing something useful.