Forum menu
The route I took was
- BBC NEws : https://www.bbc.co.uk/news/business
- Business (from the top bar menu) : https://www.bbc.co.uk/news/business
- AI Business (From the top bar sub menu) : https://www.bbc.co.uk/news/topics/cd1904ree0nt
Just did it again to get the URLS and the screen still shows this :
Just like when I use AI, I check my sources and references so I have proof of my statements
😉
People seem to think that AI should somehow be right all the time and know everything. It doesn't.
And if it doesn't there's not much point using it. I'd rather trust my hand with a spanner than a torque wrench setting from AI. You don't know where it's got it from or if it's just bullshitting/inventing.
You don't know where it's got it from or if it's just bullshitting/inventing.
isn't that the same as asking any expert though?
Just because a guy has been working in a garage for 20 years and is a mechanic through and through, it doesn't actually mean he knows the real things which are happening with your real situation at the moment. I'm not quite sure why you're so hostile towards AI well, it has many the same faults as any other information source.
It's a bit like asking a question on SingleTrackWorld. You will get some amazingly accurate answers to your questions and some incredible insight from some real experts. You will also get some complete twaddle that might kill you! You just have to guess which is which.
I'm surprised by that. I've searched "deco mesh randomly assigning mac address" and Google AI mode has replied:
"It is highly likely that your TP-Link Deco mesh system is not randomly assigning MAC addresses itself, but rather, your connected devices (phones, tablets, laptops) are using a privacy feature called Private Address or Randomized MAC Address.
Modern operating systems (iOS 14+, Android 10+, Windows 10+) use this feature to protect user privacy by hiding the device's true, unique MAC address, causing it to appear as a new or unknown device to your Deco network."
Seems pretty good to me?
(off topic but) I since worked it out (no thanks to Google, 'AI' or otherwise). In order to help with fast roaming, the Decos have "smart DHCP", even in access point mode (when they shouldn't be issuing IP addresses); this involves spoofing a new MAC address for the connecting device. Dunno why, haven't been able to find reference to it online; and can't find the search query I tried that led to "MAC addresses will be randomly assigned".
Either way, regular search is IME more reliable because one at least has the sources, and no assumption that this is an answer as such, just links to relevant context.
isn't that the same as asking any expert though?
In most of my cases it's better. Work-wise, I've spent years Google searching techie queries: Open this result in new window, nope that's not the answer, back to results, open another, nope not the exact question I was asking etc etc.
Now just one source, get an answer, a few discussion points and it invariably trawls out the right answer. I know many others use it the same way.
"Dangerous" is laughable. I love that the danger, before it actually came along, was ooh scary it's gonna kill us all and take over the world... now it's ARRGH it might tell me the wrong torque to do up a bolt! HIDE! 🤣
What I don't like about AI, WCA ?
First what I do like: People who think, preferably for themselves. People who get out there and do things. People. I like being sociable, exchanging ideas, projects, experiences, goals, ambitions, successes, failures. I like people with inquisitive enquiring minds - like you. Art and selling it on the street, messing with a car then taking it to a meet
I see IA as another step in impersonalising the world we live in, a reduction in the want and need for social contact, the need to exchange. No point trying to help people on STW, just tell them (how) to use AI like you were trying to do with me up the page - you don't even need to undu the steering arm on a 205, there enough flex in the suspension rubbers to get the drive shaft out on full lock.
Then there are things I don't like:
Thought control of the type that leads to fascist or extreme left dictators running the show and making millions miserable. I see AI as a great tool for them. Feed them the shite you want, twist the algorithms to suit your agenda. Stuff the algorithms present me on my feeds appalls me.
KIds, jeez, Google and browsers were not good for their development and AI takes away any need to think or work stuff out for themselves. They haven't got a clue what they're handing in, the 'puter did it. The return on investment for teachers is falling. And it'll be the same in the jobs. Low attention span, distracted, apathetic, demotivated, disinterested, flitting from one thing to another. Yet another discussion among Madame's colleagues on how to get them to do the work rather than let the computer do it then get anti-social when they can't do it class because they've learned nothing.
The fact I can't find what I'm looking for, this isn't just an AI problem, the browsers just don't find stuff these days. I don't know how the algorithms have changed or whether it's incresingly a needle in a haystack as the amount of crap on the Web increases but useful stuff is being both buried and now ignored if AI doesn't want it.
It's putting junior out of one of his jobs. He's a composer - producer and the amount of new content produced by AI dwarfs and drowns what real people are producing. Just as well he's also a ski instructor, though the energy wasted and CO2 produced by AI and crypto is going to put him out of that job earlier.
Drowning in a sea of useless information. Produing a generation of empty heads. Stealing jobs, eliminating sections of the economy.
My objective advice for younger people, do something a computer can't do.
There isn't a thing in your post I disagree with 🙂
This is a similar argument he Luddites had. This is not an insult, I am a big fan of what the Luddites were trying to achieve - improve working conditions, safety for workers, social justice, etc - but the press got hold of them and destroyed their reputation. Bits like "Be nice to people" seems a good idea, but the press has turned it into an evil insult.
It is always a challenge when new tools or technologies come along. They can do things, but we have to decide how we use them. Is it bad to use a lathe to make the wooden spokes for a cartwheel when you could whittle them by hand? Is it wrong to use metal spokes for a wheel, rather than the wooden ones, with all the extra effort, materials, and process that uses? Is it wrong to stop using spokes in wheels and just use alloy castings?
The same basic argument could be used for any technological advance. The key difference here is rather than improving a physical process, making something more easy, or making something better, is replacing thought process. This can lead to the situation where people no longer feel that they need to think and just rely on the AI to reproduce the thought process.
I think that this is a danger you're worried about, and it does concern me too. What I would say is there are situations where it is a useful tool, but that's not every situation. So we need to be careful to use it where it is useful and not where it's not. Unfortunately, doing the virtuous thing is not necessarily what most people do. They'll do the easy thing and just ask ChatGPT
Ask ChatGPT how many States end in the letter O. Last time I tried it told me 1. The actual answer is 4. I asked it why it gave me the wrong answer and it told me that it thought I would want the fastest answer not necessarily the correct answer. If people didn't know the actual answer they would now think the wrong answer is right. That question is really basic which is worrying.
"how many States end in the letter O"
If you asked me, I would have said three. What's the fourth one
- Colorado
- Idaho
- New Mexico
- Ohio
I missed New Mexico but I would have answered quickly which is the main thing 🙂