Remove this Banner Ad

Opinion Ai & The Future

🥰 Love BigFooty? Join now for free.

PowerBaz

Brownlow Medallist
Jun 13, 2014
22,764
32,172
AFL Club
Port Adelaide
Figured this had nowhere to go in existing threads but one I’m watching in relation to my job - as should we all.

This is the best article I’ve read and worth a read, still trying to work out are most people asleep at the wheel with this stuff or am I overthinking it.

Articles like this seem pretty compelling.

 

Log in to remove this Banner Ad

AI has some benefits. Might be close to what Ange is truly thinking.




Can y'all imagine the chaos when A.I. begins doing this perfectly? It's not far from it. We won't be able to trust any video anymore. We'll really need to improve our interpretation skills to discern what is true, what is credible, and what is false. (It's not good at the moment, to be fair.)

In a way, we'll be back to how it was prior to photograph and video. Only the scale will be astronomically larger than it was. Gossip won't be restricted to a small area, but able to reach the whole world.
 
Plus it is already having an effect on vulnerable people’s mental health. In the US some people are already believing that actual angels or demons are talking to them through ChatGPT
 
Plus it is already having an effect on vulnerable people’s mental health. In the US some people are already believing that actual angels or demons are talking to them through ChatGPT
Interestingly though I've heard many accounts of the opposite. That a lot of people out there basically use it for therapy with really good results.

I don't know what's true either way, just random stuff I've seen on Reddit.

I know its great for my mental health when it turns a task that would have taken me an agonising two days in to a 5 minute task :)
 
Interestingly though I've heard many accounts of the opposite. That a lot of people out there basically use it for therapy with really good results.

I don't know what's true either way, just random stuff I've seen on Reddit.

I know its great for my mental health when it turns a task that would have taken me an agonising two days in to a 5 minute task :)
I accept that and I use it myself for certain things but when you mix it with poor mental health, to replace actual relationships or extreme religiosity it is a loose cannon.
We have all seen how Q radicalised the US.
 
Amazon planning to reduce its corporate workforce as it plans for AI to take over a lot of their functions in coming years. Also increasing the use of robots to replace some other workers.
Will make it difficult for those in the US being told that with the Big Beautiful Bill slashing Medicaid that their solution to get or retain medical cover is to “get a job”…. in a potential shrinking job market.

 
Last edited:

Remove this Banner Ad

Always funny reading this stuff as a software dev and then actually trying to use these tools for work. So overhyped. They'll get there one day but they need to make some big breakthroughs first.
Yeah funnily I think the tools are actually way better for science and low level analyst types... it gives you an access point to so many different languages. Has been revolutionary/life changing in a career sense for me.

Would totally get why its way more meh for those that actually live and breathe IT.
 
Yeah funnily I think the tools are actually way better for science and low level analyst types... it gives you an access point to so many different languages. Has been revolutionary/life changing in a career sense for me.

Would totally get why its way more meh for those that actually live and breathe IT.
It's basically replaced google searches, combing through docs and stack overflow which is no small thing. I just haven't found it very good at doing the actual work and they seem particularly bad at debugging especially if it's not an obvious bug.

I think most people who use LLM for their area of expertise quickly realise how much bullshit they produce and that's just a product of using statistics to predict the next token. Feels like they are a lot further off fixing that issue then the tech bros are willing to admit. Again that's not to say LLM aren't useful.

Big fan of Yann LeCun who's the Chief AI Scientist at Meta. He has a much more balanced opinion on the current state of AI and LLMs.

 
It's basically replaced google searches, combing through docs and stack overflow which is no small thing. I just haven't found it very good at doing the actual work and they seem particularly bad at debugging especially if it's not an obvious bug.

I think most people who use LLM for their area of expertise quickly realise how much bullshit they produce and that's just a product of using statistics to predict the next token. Feels like they are a lot further off fixing that issue then the tech bros are willing to admit. Again that's not to say LLM aren't useful.

Big fan of Yann LeCun who's the Chief AI Scientist at Meta. He has a much more balanced opinion on the current state of AI and LLMs.

I think its underselling it a bit.

If I have a pdf of various things and properties and want to convert it to say a python dictionary in a particular format you cant do that with stack overflow.

Also code from search requires full customisation once downloaded, whereas chatgpt can also do this step a lot of the time.

Its intelligence and autonomy is way overblown but its consistent ability to complete repetitive tasks and its ability to quickly take you from say a 2/10 to a 5/10 in almost any language or program is incredible.

To me it sounds like you're much more specialised and it will have less application for you for that reason. I use it for hours most days.
 
Last edited:

Remove this Banner Ad

Opinion Ai & The Future

🥰 Love BigFooty? Join now for free.

Back
Top