I don't understand the point of AI

Firstly I think we have to decide on what type of AI we are talking about - it is so broad nowadays and yet most people think of generative AI tools such as ChatGPT, but we also have AI responsible for online algorithms (YouTube recommendations, social media feeds, etc.), AI for video game enemies, etc. etc.

From reading this thread it's clear that most of us are talking about the former: ChatGPT and its competitors. Personally, I can see the benefits in terms of increasing workplace productivity when it comes to certain tasks, but I'm against it in the vast majority of scenarios. I think the most dangerous thing is the way it can so quickly hallucinate information and feed it back to you as if it is fact, meaning that if you engage in a conversation on a topic you know nothing about, the AI can easily mislead you into thinking something completely false. I'd recommend taking a subject you're knowledgeable on, then asking AI very specific questions about it and see how quickly it goes down a rabbit hole of lies, and then you'll see just how "intelligent" it is. For example, I asked AI to explain ARP (Address Resolution Protocol) to me, and it got it absolutely backwards.

However, I feel like even in the cases where it can provide useful information it's still a bad thing. Let's say you use it to brainstorm ideas, it often does a decent job and it can help inspire you with new ideas, but if you're not careful you'll go to it for every single creative task and before you know it you have no imagination of your own.

It can certainly be advantageous, but I feel it'll mostly be advantageous to the wrong types of people: hackers using AI to crack passwords, falsified information being used to push propaganda campaigns, AI chat agents being utilised by scammers, etc.

I know you aksed for people to outline the advantages, and I have been super negative, but even when I try to look at it positively all I see are flaws. Sorry haha.

I hope I don’t become unconscious with my eyes open! I like the ease of use of the iPhone & other Apple products, also they are more secure than others. 

Bunny said:

And if someone tried to physically hold my eyes open, I expect the FaceID identification would fail anyway, as their hands would be obscuring too much of my face. 

Yes, and I think there are all sorts of other negative things that can happen in life and that I can get carried away if I’m not careful. I often tell myself to look at the worst possible scenario (regarding AI - the benefits or downsides) if muggers  accessed my phone.  In all likelihood, I would be so relieved to survive such an attack that the stolen and possibly accessed phone wouldn’t be my prime concern. Sometimes it puts things in perspective. Strangely, I have tried using ‘the worst possible scenario technique’ for things like doctor’s appointments, meeting people and shopping, but for those sorts of things it doesn’t work.

ArchaeC said:

I like the ease of use of facial recognition but like you, I am concerned about using it being used if I am mugged or unconscious.

I don't know about others, but Apple's system has a feature that addresses risks like these:

"Face ID recognises if your eyes are open and your attention is directed towards the device. This makes it more difficult for someone to unlock your device without your knowledge (such as when you are sleeping)."

I think that if I were mugged, I'd be just as likely - upon being coerced - to give up my passcode, as I would be to comply with a demand to look at the device with my eyes open.

And if someone tried to physically hold my eyes open, I expect the FaceID identification would fail anyway, as their hands would be obscuring too much of my face. 

I am siding with Sir Tim Berners-Lee on this AI front (he was referring to the operation of social media platforms and how they have been deploying AI algorithms):

"It is really important that we realise this toxicity comes from the algorithms and you can change that," said Berners-Lee. 

"There's some coder who can just sit there and tweak the way the AI is trained so that it can be more healthy, constructive, creative and helpful".

https://www.euronews.com/next/2024/11/21/world-wide-web-inventor-says-algorithms-should-be-blamed-for-toxicity-on-social-media#:~:text=in%20youth%20culture.-,%22It%20is%20really%20important%20that%20we%20realise%20this%20toxicity%20comes,constructive%2C%20creative%20and%20helpful%22.

Back in my first career, I used to work in the large IT department of a multi-national company which used to consult with Sir Tim (when he was "Tim") on e.g. the (very) early adoption of: the internet, email and the formative stages of associated application development (all very much beta). 

You know, when people had role titles like "Systems Engineer" (with the emphasis on Engineer) and a business continuity mindset (across all in the department) meant the luxury afforded us of planning and testing wasn't a euphemism for ..."disaster" already happened and our customers noticed it first (because we were forced by non-IT Line Management bean-counters to try it first in live production) and "recovery" was an automated computer job control language / procedure alternate schedule suite of production-preservation actions (with 98.99% availability of all services with little human endeavour) which we would weather-eye (post mortem if necessary) the next morning over our first coffee ...rather than an interminable flurry of social media posts along the lines of "we are working on it, now customers have reported it" (with the bean-counters in a super-tail-spin-lather by that inevitable stage!).  The modern IT professionals know what ought to be done and have the competence / capabilities ...but are often not allowed to do so ...hello, AI World (if we don't catch a good firm grasp on the tail of the dragon).

My own outlook on AI is "harness it, but don't accept being hassled by it".  Design it, test it, review it and fix it if it is an annoying burden.  What I don't subscribe to is the: "ooh, new sparkly thing ...must use it to show off progress and scrap jobs".  Like sny other IT "thing", it should become a stable tool for business operation and providing products and services actually welcomed by members of the public (not allowed to be the wild west play thing preserve of a handful of tech-bro-oligarchs with a, let us say, less than altruistic moral compass).

I like your imagery of "AI as an Ice Breaker".  I shall ponder that further.

All this trip down memory lane means over breakfast this morning; I will be reflecting upon a happy few weeks of yesteryear: spent booking constrained minutes of satellite time slots via the Goonhilly Earth Station satellite dish (before trans-atlantic undersea internet cables were a commercially viable option). 

Why? 

Because unsavoury characters "accidentally" dragging their ships anchors across international undersea cables, these days, means all of our resilience hopes may indeed rest once again upon the bright young things of Cornwall et al. being "A1" (A one) not "AI" (Artificial Intelligence). 

For the benefit of younger viewers: "A1" used to be the commonplace shorthand for "excellent" (very good or well, excellent / derived from a vessel in Lloyd's Register of Shipping having been designated as equipped to the highest standard; first class). 

Our New World Order means many professionals and populations alike; are going to need to have a good lateral brainstorming session, step up, and link innovative capabilities to demonstrate and preserve critical infrastructure resilience in the face of hostile actors.

Some of our Nordic cousins never stood down their resilience mindset and preparedness training (we, each of us, need to dust off and spin-up our own versions of "resilience first"). 

As much as anything; it is about diligent, relentless, resourcefulness.  Being willing to be flexible and creative about problem-solving and getting on with implementing improvements in support of resilience prevailing.

The sands of time mean; it is no longer about the luxury of indulging in:

- AI nightmare-mongering and anxiousness.

Rather:

- asking how AI helps in bolstering resilience is the watch phrase of today.

There is no more bandwidth available for accommodation of political rhetoric, or grand standing, or point scoring, or finger-pointing, or name-calling, or prevarication.  Leaders must rise to meet the challenge.  Populations must play their individual parts too.

Now, where do we feel AI can be reliably harnessed in support of this endeavour? 

(As opposed to enduring the hassle and the immature backdoor security exposure ...to potentially be exploited by the subterfuge of the unwelcome actors?).

AI needs to shape up ...or ship out. 

(The UK's history as a maritime nation infiltrates our vernacular ...pervasively ... I couldn't resist that one!  That is to say: "used as an ultimatum to someone to improve their performance, or behaviour, or face being made to leave").

I believe we are all going to be too busy, for sometime, (dealing with the essentials of resilience demands), to spend too much time polishing a promising laboratory experiment - which is demonstrably still in its infancy. 

We can still look forward to the day when AI's training wheels may be (safely) discarded.

I run a team of younger technicians and would hear them make comments about AI.  I avoided chat gpt and what not.  I am not afraid of it and somewhat amused by the concept.    I had two engineers fly in to run an intense class for my techs and first thing they did was pull up AI and put crazy questions in there.  Everyone had a good laugh and it was a great ice breaker.  I went home and pulled up an AI tool and asked it if was cable of being Jacis for me as in Iron Man.  The responses were pointed and disappointing.  Still to this day I would be on board if I can walk around and have AI be Jarvis and makes things more efficient.  I don't want it to do my work for me but be at least good enough to carry the tools I need and reduce my effort to input my queries and get a unbiased real deal output  otherwise AI is an overprice icebreaker for the beginning of a grueling class...

I am firmly in your corner too Cloud44 / Catwoman.

I also find lack of manned checkouts frustrating, I don't like the self service, I find it all stressful.

TheCatWoman - Based on this and the other conversation about modern tech, I feel like we are on exactly the same page with our thinking. It's a relief to know there's someone else out there thinking the same way. As I keep saying, I just want a simple life, I feel everything has got way too complicated and out of control.

I like the ease of use of facial recognition but like you, I am concerned about using it being used if I am mugged or unconscious. I make sure my banking apps & some others need extra information to log in, but I will continue to use facial recognition for the phone as I would forget a PIN or password.

I don’t particularly like AI, but I see it’s potential for things related to health, climate change mitigation, criminal investigations and so on. Holocaust survivors have entrusted their memories to AI to ensure their stories continue to be told down the generations. 

I have used AI to write personalised poetry. I enjoyed inputting the relevant information and WhatsApping the finished poem, complete with personalised AI picture. Having done that a few times, I don’t feel the need to do it again, but it was fun at the time. I hope that AI wouldn’t deter potential future poets, as the AI versions can’t match the literary depth of human endeavours.

I think the potential for abuse of AI is rather frightening and the difference between fact and fiction could become increasingly blurred as technology becomes more sophisticated. In many ways, I would rather that we didn’t have AI, but now that we do, we just have to get on with it, as most of us are already using it, even if we don’t realise.

AI can be many different things. It can help us, or It can cause problems and even discriminate against us if it's not programmed correctly.

I like the AI in video games which enable me to play against it without other humans, and the AI on kindle that suggests other books I might like as it's helped me find many books I didn't know about and have enjoyed. 

I don't like AI facial or fingerprint recognition on things like phones - if you got mugged, the perpetrators could just hold up your phone to your face or put your finger on the screen to access everything on it. I also don't want my household appliances making their own decisions - I am capable of deciding what programme to wash things on and checking whether I've run out of something.

I think there are a lot of potential uses in health & social care such as reading scans and having robot carers and companions.

AI can be used to do things like judge whether someone who has been arrested should be given bail, or for assessing benefits claims or allocating social housing. But there is a danger of discrimination with these tasks, so the machine learning has to be handled very carefully. And if course, self driving cars need very specific programming for safety reasons.

I see A.I. in a similar light to the calculator that so changed our lives in the early 1970's....

OR the GPS that has replaced mapreading skills, adn replaed them with greater effciency (mostly..). 

I was being scammed into my first foray on X so I used their A.I. (called Grok) to get me the data I needed to evaluate my situation correctly.

Interesting stuff. It won't replace God for me.though.

Bunny the uses you've outlined fill me with horror, I want to choose how long I wash my clothes for, how much detergent to use and how hot the wash is. It's bad enough already with so many machines not having a hot wash option, and so many machines want you to use specific types of detergent, nearly all of which I'm allergic too, I think swapping more detergents for less electricity is a false economy too, does no one take into account the amount of energy used in the creation of laundry products.

I also don't want my fridge telling me what I should buy and least of all do I want it oredering my shopping for me. It might be great for some people with particular disabilities. I don't want biometric phones and stuff, what happens to all this data? Will my time be owned by a multinational company like apple?

I think that pushing automated systems at people with disabilities is ultimately lazy, it dosen't encourage real contact and people will end up having to accept stuff they don't want or be seen as being awkward and theres enough of that already.

Choice about all this bothers me too, we are increasingly having less and less choice and certainly the choice to not use these things is decreasing. No maps of places because everyone can use their phones to find things, having to pay for thing and order them via an app, even when in store! If I go to Llanduddno, which I try and avoid as much as possible, its like being taken to another planet and it's only 50 miles up the road! So many people who live here don't go there, despite it being where all the shops are, because they've no idea how to navigate the place, M&S cafe feels like the Starship Enterprise! Then you get the opposite of people comming here from other places who can't cope because they have to do things manually and theres no uber.

In specific respect of autism, this recent paper, for example, explains that:

"results reveal an early yet promising interest in integrating AI into autism assistive technologies. Exciting developments are currently underway at the intersection of AI and robotics, as well as in the creation of wearable automated devices like smart glasses.

These innovations offer substantial potential for enhancing communication, interaction, and social engagement for individuals with autism."

Breaking Barriers - The Intersection of AI and Assistive Technology in Autism Care: A Narrative Review

It seems to be good for looking at things like mammograms, scans and xrays to aid diagnosis, it's been tested and seems to be slightly better than humans at diagnosing cancers at early stages.

But on the whole no I don't like it, it's not actually intelligent either, it does the task it's asked to do very fast, using a range of sources to gain information, but it cannot actually think. I don't like having an Ai description of a search I've made, it appears at the top of the screen above the sponsered links, so it takes even longer before can even try and find the information I asked for. I think as it becomes more and more widespread,  all most of us will experience is a whole other realm of "computer says no", I think it will become ever more embedded in our lives and our smart phones, which will be needed to do more and more ordinary things, taking away agency and power over our lives.

AI gives bad actors a big advantage in censoring what we can see, find out and learn, of course, there's not just concious censorship, but unconcious bias and forgetting about information points, and how others use things and the physical differences. Like here, when our phones can only pick up a signal from ireland, or that we have 3,000ft of slate sticking up out the ground that interupts signals. If AI translates a book for us, how will it cope with the nuances of language, translation isn't just a word for word copy, it has to make sense and it has to convey the sense the author does in the original.

It all seems geared to taking choices away from us "to make our lives easier", but easier how? Will it give us more time to veg out in front of the telly, to buy more rubbish or be influenced by people or AI avatars? I don't really have a problem with making choices, although some people think I do because I say no too much for thier liking, but I'm a fairly simple soul who dosen't need clutter and cargo, nor do I see the need to make simple binary choices as some sort of game.

Cloud44 said:

Can anyone see the advantages? Rather than be dismissive of AI and shut it down, I'd like to learn from others and find out what they get out of it and maybe find something I am missing, or how it could benefit me.

AI can benefit us in so many ways that I'd struggle to offer any kind of reply that does them justice.

Just a couple of examples of everyday use at home include:

• How biometric security features (such as Apple's facial recognition system) provide convenient protection to phones and other devices.
• How my washing machine automatically works out and delivers the right detergent doses and washing times for each load.

In the healthcare sector, the scope for benefits is enormous; this recent press release explains how it can support independent living, for example:

GOV.UK - AI sensors on fridges and kettles helping vulnerable people to live independently

Other healthcare benefits can be found in screening (including analysing imaging / scan results). This project, for example, "has the potential to save more than 11,000 lives and save the NHS at least £660 million over 5 years" by providing clinical grade testing at home for people with certain conditions or risk factors:

NHS England - Healthy.io: Smartphone albuminuria urine self-testing

On a more personal level, I find it interesting - and sometimes helpful - to talk with AI chat bots (including for support with my mental health conditions).

I actively dislike it. It seems to do nothing but suggest things that are irritating and try to railroad you down paths you have been on before, but might not want to repeat.