Is AI friend or foe in the Clone Wars?

Status
Not open for further replies.
I am curious to know the process behind how it comes to the conclusion about which details are accurate. Does it know to search reliable knife dealers for true facts and photos or does it just use most frequent facts and photos on the internet. If someone made a millions Spyderco PM 2 clones and had more photos and fake facts on the internet about them would AI start calling real PM 2s clones.

It is an interesting rabbit hole to go down. Still, I was impressed with how much accurate checking it was able to do from a few eBay photos.
 
I don't know how useful it is currently for authentication of knives, but the conversation has already broadened, so I'll add a couple points.

I lead a small academic research group and find that use of AI in a variety of contexts is exploding.

It's easy to ask ChatGPT a simple technical question and receive the wrong answer. It will waffle around in an unreassuring way if you correct it.

I've also received AI generated draft manuscripts from students with ESL, who usually write poorly. At a glance the writing is superficially much better than their usual, but the flow of ideas is all over the place and it's essentially slop that has to be completely rewritten.

However, we recently explored a commercial level access to Claude AI agent, and obtained amazing results in a carefully controlled context.
My student graduated this year with his PhD after 5 years of work which culminated in a new scientific software package. In the span of just 2 weeks, working with Claude, he was able to completely rewrite the software, with significant improvements. He fed it his entire thesis, his code, all of his papers and references. He had many guardrails, benchmarks, properties of the solutions etc to verify, and so he was able to confirm that Claude's version is accurate and better.

Our sense of the current state is that used as an assistant for specific tedious tasks, it can be amazing. Given uncontrolled latitude it will produce nonsense. It's already too powerful to ignore.
 
marcus aurelius said:
Believing AI is a lot like believing the news. IF you've ever been close to a news story and know the real details, you notice just how wrong the news gets it. AI is similar in that the more you know about a topic, the dumber AI seems. If you know nothing on a topic, then AI seems really smart.

So when it comes to knives, I'd trust the people on here to come up with the correct conclusion long before AI even got in the ballpark.
Click to expand...
Same. People on here actually have and use those same knives I wanna know about.

Besides, exploration is part of the joy about this hobby.

You come across things you wouldn't otherwise see, or someone gives you suggestion that you'd never think of, someone mentions a knife you'd never think of or you never heard of...

AI is like laser precision into - basically nothing. It's designed to tell you what you want to hear. So you are talking to yourself...
 
It just seems to me that so many people want easy peasy and not use the intelligence they have to become more intelligent. I think it’s a bad deal for the future of humanity to rely on AI to do the intelligence and work to think with our brain and not just feelings. I believe it will only serve to dumb down society even further than has already happened. Exercising our mind is to our advantage. I know sometimes we can get overwhelmed or maybe intellectually lazy and desire some help but that’s why we have books and have the ability to read and write and study at our own pace. I also have seen several times where AI got it wrong or flawed so that leads to backwards thinking, misinformation and confusion. What more could go wrong if people using AI weren’t smart enough to figure it out and know that something is wrong ??? There are many other things that can be destructive because of AI, like water usage that leaves us without enough fresh water, energy shortages and a ripple effect throughout commerce, energy and the environment. Maybe in the short term some can see a benefit but I believe it will be temporary and then many will realize it is to late to correct course. This is my opinion.
 
Lesknife said:
It just seems to me that so many people want easy peasy and not use the intelligence they have to become more intelligent. I think it’s a bad deal for the future of humanity to rely on AI to do the intelligence and work to think with our brain and not just feelings. I believe it will only serve to dumb down society even further than has already happened. Exercising our mind is to our advantage. I know sometimes we can get overwhelmed or maybe intellectually lazy and desire some help but that’s why we have books and have the ability to read and write and study at our own pace. I also have seen several times where AI got it wrong or flawed so that leads to backwards thinking, misinformation and confusion. What more could go wrong if people using AI weren’t smart enough to figure it out and know that something is wrong ??? There are many other things that can be destructive because of AI, like water usage that leaves us without enough fresh water, energy shortages and a ripple effect throughout commerce, energy and the environment. Maybe in the short term some can see a benefit but I believe it will be temporary and then many will realize it is to late to correct course. This is my opinion.
Click to expand...

Reads suspiciously like AI

☺️
 
I’ve messed around with AI just to see what all the fuss was about and found it to be mostly unreliable. People think of it as almost a living, reasoning, thinking thing when all it really is is a dispensary of information from various sources - some not very reliable. So it doesn’t mean the info you get from it is correct. It can only spit out what it’s programmed to spit out - right, wrong or indifferent.

Someone already said this in a prior post and it’s true. If you happen to know it’s wrong about something and call it out, it’ll backpedal and tell you you’re right. Sometimes it’ll try to convince you you’re wrong first, but if you confront it with more facts it’ll cave and say you were right.

Almost as though it’s been programmed to gaslight the uninformed into believing certain things. . . Hmmm, I wonder why it was designed to do that?

Edited to add that I’ve become pretty good at spotting AI slop narration on YouTube videos usually within the first minute. The speech pattern and cadence is the tell. It all sounds the same.
 
Blues said:
I only use AI for confirmation bias...so that I can say to myself..."See, you were right all along".

:cool:
Click to expand...
That is a little like how I rewatch Nick Shabazz videos for knives that I already own.😁

I think rules, regulations, education and common sense when it comes to AI is all a good thing.

The task it is being used for is also important. I don’t want AI to be the detective, judge, jury and executioner. I don’t want it to decide weather or not my kid gets into a college. I think that particular ship has already sailed.

I do like it for picking out a decent chocolate chip cookie recipe. It is when it tells me to add 1 cup of antifreeze because of how I voted in the last election we will have serious problems
 
Another thing to remember is that "AI" is not one monolithic thing. There are many, many LLM models out there, all are a little to a lot different. Even the same AI company will have multiple models. Most people read a couple headlines and think they got a handle on AI. You have to do some digging on your own. Every time I do, I learn something new.
 
Alright, since we've pretty much moved on completely from AI and knives...I think we'll put this one on ice.
 
  • Like
Reactions: Hig
Status
Not open for further replies.
Back
Top