Le aziende del Regno Unito sprecano milioni in strumenti AI in stile ChatGPT
di tylerthe-theatre
Le aziende del Regno Unito sprecano milioni in strumenti AI in stile ChatGPT
di tylerthe-theatre
13 commenti
My experience when consulting was that businesses are figuratively putting the cart before the horse.
When else would you ask “how can we leverage this tool” after buying it?
AI not a replacement for years of experience and training then? Im shocked. Maybe planning those fat exec bonuses from laying off staff was a bit premature. Mind you half the dinosaurs in the C-Suites can barely spell AI…
You have to be in the game, if someone figures it out and you haven’t been playing then you are going to be so far behind.
The problem with AI reporting is there’s been no neutral reporting.
It maybe waste now, but if the companies have re-organised their data to be much more machine friendly then that’s future proofing
The problem we have is we are not discussing how lean businesses can get from ML enabled workflows and how many actual humans they are going to need, it may not happen soon but that’s the path we are now on
Presumably because so many of them are just a wrapper around a £20/month ChatGPT subscription…
It’s amazing how much money is being spent investing in AI wrappers.
Where i work i wouldn’t say its a waste. And the development hasn’t costed much. Team at India did the development and a lot of the coding etc is sourced from Microsoft m, its not a exclusive AI tool made from ground up. Its just a layer added on tip of the existing AI tools that are mainstream.
Its helped with quite a lot of work from my analysis, its more of a glorified search engine and not a replacement. But that still saves a lot of time and helps the business not have to hire more people
The problem is many of these businesses have no idea what or how to use AI and they’re just being led by the headlines
Good. Hopefully, they learn from this and remove the dogshit from their websites, etc.
Some of the ways AI is being advertised is just bizarre anyway. There’s one advert where this person asks the AI whether some flowers would go with other flowers, which is completely strange because that’s down to the individual to discern, and asking an AI something that can be clearly seen feels daft.
There was also a story of a company considering putting ChatGPT into dolls for kids to engage in meaningful conversations. What happened to imagination, or talking with mum or dad? The ideas are basically throwing AI on anything which may have some link, without actually thinking of how to best utilise them.
I’ve yet to even see an enterprise purpose for it that is actually justified.
* Providing answers to questions – Google, and actually read + comprehend the relevant information
* Automating tasks, e.g. changing from 1 format to another, extracting certain info from documents, etc – If it’s something that is consistent enough that it can be safely delegated to chatgpt then it can be scripted or done with a unix command.
* Image generation – The only actual novel use, but there are a lot of ethical concerns, and most places don’t actually need to churn out AI slop
My sister’s boyfriend always raves about how useful chatgpt is and how it’s magical for automating his work, but the things he talks about are trivial to use something like python for. He does say that he loves it regardless so he’d probably still have a subscription even if he knew how to use python, but for a corporate environment most places will have an IT department already, delegate the task to them and you’ll save so much.
Take a step back and it’s one of those situations you see in lots of industries.
New tech can lead to first mover advantage, but also first mover *dis*advantage. Especially if you’re stuck with an expensive tech and data stack which is hard to decommission and concurrently pivot to a new one.
Some companies are in FOMO mode. Some are going all out. Some are dipping their toes in it. Some are waiting a bit.
Big picture we’re also going to be in an AI bubble. Same as the Dot Com bubble. But out of that you got a few tech giants which grew up big – and fundamentally has improved consumer lives a little bit. It’s just Joseph Schumpeter’s capitalist “creative destruction” playing out.
I think AI is one of those things that will probably help businesses overall, but as an assistant like they used to be, not the replacement for staff all the dumb, short sighted businesses owners who were rubbing their hands together at the idea of sacking all their staff thought they were going to be.
They integrated AI into one of our workflows, it seems like a waste in that it just converts your message into an email without you having to type as much or it’s used to search internal databases.. that’s it.
And they probably spent millions upon millions on it and held bit C-Suite meetings glazing each other on “workflow optimisation” “being ahead of the curve”
“At the forefront of AI optimisation in the industry” etc etc
Oh, it’s that MIT NANDA report again. Keep up, CityAM. That came out in July.
And though it satisfies my confirmation bias, it’s a pretty crap report where the researchers were trying to talk up their own output but saw the conclusions run away from them.
Unfortunately, right now, there isn’t good evidence for the waste, though there are plenty of anecdotes.