I only use Gemini on PC thus far but it definitely remembers conversations. It shows an abbreviated title list on LHS so you can easily spot & enter that conversation again & carry on. Recently it introduced a different looking 'chat box' area, forgot the name. Apparently its something like incognito mode where it purges after some time amount & supposedly does not use those questions for learning. And of course I trusted it completely when I asked 'where would be a great spot to bury my 24kg of gold bars?'. I should use ChatGPT more just to toggle back & forth. I had a good session going on some engineering topic, it was charting graphs, then it suddenly ran out of gas & left me high & dry.
-
Scam Alert. Members are reminded to NOT send money to buy anything. Don't buy things remote and have it shipped - go get it yourself, pay in person, and take your equipment with you. Scammers have burned people on this forum. Urgency, secrecy, excuses, selling for friend, newish members, FUD, are RED FLAGS. A video conference call is not adequate assurance. Face to face interactions are required. Please report suspicions to the forum admins. Stay Safe - anyone can get scammed.
Anyone using ChatGTP?
- Thread starter Mcgyver
- Start date
I've heard Alexa called narrow or weak AI. Not AGI general intelligence. She is basically a nosey snitch to her BFF's in the Cloud. No Alexa for me.
Sure it can. Go to menu and search for past chats or use the storybook function which has all your chats.
See above. I have no trouble picking up where I left off.
Ya, searches are waaaay more powerful now. I find that how you ask the question matters.
I bet it does. We just have to figure out how.
Ain't that the truth!
I agree.
I confess that I started out VERY suspicious and very negative. I still have concerns but they are for much deeper issues than my usage. For my own usage, I have completely flipped to the point where I almost love it. I'm not gunna let concerns about how much it knows about me get between me and my ability to enjoy using it.
If you like Gemini and ChatGPT, try perplexity. It's my current favorite.
And here is a tip - next time you think you don't like something about it, ask it how to fix it....
And another tip. The pro versions have more features. But.... You can ask it how to get a free upgrade and it will help you do that! Amazing really!
afaik, there is no such thing as AGI, in fact it's far from decided as whether it is even possible .... i.e. define unconscious. the other idea that to me makes sense is that you can't there from here. If AGI is possible, it will likely not be AI with more horsepower. AI as we have it is a probability engine (with lots of training) - it only seems like it's thinking because our brains (mis) interpret patterns as intelligence.
As i understand it, where it will improve, is with longer persistent memory, 10 million tokens instead of 10 thousand This doesn't mean memory like storage, but memory specific to the theme/train of thought. Still, that would be just a better probability engine - it is still not thinking, planning, engineering, working toward a goal etc.
Remember Eliza? 200 lines of code in '60's that had some convinced the computer had become a sentient being. It would trigger on keywords and ask questions like "how's your mom?" and people thought it intelligent. Headlines like "seeming conscious" or it "understands human emotion" appeared. how easily the masses are misled.
Attributing "thought" to current AI is no different, just that, if you add in the training, there are millions of lines of code. I also think there is currently a mania like hype around it, that like the reaction to Eliza is BS. e.g. Tempur Pedic mattresses advertise their mattresses as having AI ..... the world has gone insane and brains turned to mush.
The 1960s Chatbot That Convinced People It Was Human: How ELIZA Broke Brains Before AI Was Cool
Photo by Denis Cherkashin on Unsplash
Last edited:
DPittman
Ultra Member
Interesting.i guess it was the way I was using Gemini maybe. I just use button on the side of my phone and a Gemini window pops up and I ask it a question and then the answer is read out to me. That window disappears and I was never able to go back to that conversation as far as I knew. Matter of fact I specifically asked Geminin if it could recall past conversations and it told me NO it has no memory!
I will now dig deeper into it. It seemed odd to me that ChatGPT was doing it but Gemini was not.
Yes I have A.I. technology at my fingertips but Genuine Stupidity runs deep through the rest of me.
maybe it's just some phone app lite thing. I use chat on the browser and all chats, right back to their beginning, are there on any device and browser i chose.
Where it is really good imo is open ended questions. Last night on the phone I was asking about integrating two programs (accounting and contact management). Rules, what is shared, how its shared etc .... it gets complicated. It was awesome at serving up a lot things I hadn't thought of. This morning on my desktop, the conversation is there. Where (I know) the wheels will fall of the cart is the specifics of execution.
Where it is really good imo is open ended questions. Last night on the phone I was asking about integrating two programs (accounting and contact management). Rules, what is shared, how its shared etc .... it gets complicated. It was awesome at serving up a lot things I hadn't thought of. This morning on my desktop, the conversation is there. Where (I know) the wheels will fall of the cart is the specifics of execution.
I "think" that side button or voice access thing is not really Gemini. It's just a quick access to it and is really more like Google Home.
Everyone and their mother is doing AI now so what is what totally depends on context.
I like @Mcgyver's thoughts on the matter.
But none of us know what is going on in the military labs in the US, China, and India.
As I said earlier, I am no longer afraid of how it will be used by evil. I am just enjoying my own use of it. It beats regular search tools by miles. Stuff like writing programs, image generation, and local programming is just amazing even though it has flaws.
This. In spades. AI is nothing more than language models at this point. It is light years (in computing space) away from becoming "sentient". Be very worried the day that actually happens.
Elektrishun
Ultra Member
Attributing "thought" to current AI is no different, just that, if you add in the training, there are millions of lines of code. I also think there is currently a mania like hype around it, that like the reaction to Eliza is BS. e.g. Tempur Pedic mattresses advertise their mattresses as having AI ..... the world has gone insane and brains turned to mush.
The 1960s Chatbot That Convinced People It Was Human: How ELIZA Broke Brains Before AI Was Cool
Photo by Denis Cherkashin on Unsplashmedium.com
Leave it to the folks trying to sell us stuff to take words and make them "buzz" to increase sales. Like "fat-free" and "sugar-free" (but what did they replace the fat and sugar with? Is it better or worse?). How about "green", "environmentally friendly", and "ethically sourced"? How about "smart technology"? Now they're using "AI" in the ads to sell a product...
Oh, and one more, "Made in Canada"🙂
Elektrishun
Ultra Member
After spinning my wheels to get someone else's tweaked Arduino code to work for a project I decided to find a smarter person to do my homework for me. This is where I enlisted ChatGPT. All I can say is that was a long frustrating road to nowhere. Sitting back and looking at the big picture I think the poor result was more a "me" problem. Likely bad data in, bad data out. If ever I try going down that road again I will attempt to form the questions more carefully and see if I get a better result. Or just do it the old fashioned way by learning and doing it myself.
140mower
Don
If you go that route, what are you going to do for the rest of the day? 🙄
Elektrishun
Ultra Member
I am retired and I still find myself working on a project problem late into the evening as though I have to be somewhere the next day. Old habits die hard and the projects will never end.🙂
One of the things I have learned as I expand my own use of AI is that they are not all equal. ChatGPT is decidedly not good at writing code. You might have done better with Claude Sonnet 3.5. The other difficulty is that they have all figured out how to make money with AI subscriptions. The more clout you want, the more it will cost you. Surprise surprise!
Lots of important thoughts here across philosophy of the mind to the political economy of mass consumption.
As it was already noted, pattern recognition is not intelligence nor wisdom. All the great art, design, and research that AI gathers (steals, uncredited) is based on previous knowledge and expertise production of both industry and academia. There will (very quickly) come a time when the majority no longer wants to draw, design, or discern (and bean counters and CEOs see fit to replace needy humans). As AI outputs are valorized and flood the market so will these outputs quickly become the next input. Data mad cow disease?
As someone else pointed out, the experts can pick out the flaws in their own domain. How many are using AI as masters of their own expertise?
I teach a new generation of youths that deploy AI in all aspect of life. There is clear difference between those who deploy the embodied learning of going through the motions and "getting their hands dirty", compared to those 20-year olds that try to keep their shirts clean reading their generated "reports".
There is a clear difference between the few and majority of kids that rely on AI who don't even know what they don't know and can't begin to evaluate the stuff coming out of the black box. They are born right into the trusting normality of all-knowig Gemini and ChatGPT. From corporate news anchor to the corporate knowledge robot.
As it was already noted, pattern recognition is not intelligence nor wisdom. All the great art, design, and research that AI gathers (steals, uncredited) is based on previous knowledge and expertise production of both industry and academia. There will (very quickly) come a time when the majority no longer wants to draw, design, or discern (and bean counters and CEOs see fit to replace needy humans). As AI outputs are valorized and flood the market so will these outputs quickly become the next input. Data mad cow disease?
As someone else pointed out, the experts can pick out the flaws in their own domain. How many are using AI as masters of their own expertise?
I teach a new generation of youths that deploy AI in all aspect of life. There is clear difference between those who deploy the embodied learning of going through the motions and "getting their hands dirty", compared to those 20-year olds that try to keep their shirts clean reading their generated "reports".
There is a clear difference between the few and majority of kids that rely on AI who don't even know what they don't know and can't begin to evaluate the stuff coming out of the black box. They are born right into the trusting normality of all-knowig Gemini and ChatGPT. From corporate news anchor to the corporate knowledge robot.
Jimmy Fallon Lost It After The CEO Of OpenAI Told Him He “Cannot Imagine” Raising A Newborn Without ChatGPT
“How embarrassing to say out loud,” one person said in response to Sam Altman’s comments, while another wrote, “We are regressing as a society.”
These tech bozos want us to depend on their products to know how to wipe our own asses. The new generation is falling in line.
I overheard someone arguing with Siri the other day, because she forgot to set his alarm clock 🙄
Jimmy Fallon Lost It After The CEO Of OpenAI Told Him He “Cannot Imagine” Raising A Newborn Without ChatGPT
“How embarrassing to say out loud,” one person said in response to Sam Altman’s comments, while another wrote, “We are regressing as a society.”www.yahoo.com
These tech bozos want us to depend on their products to know how to wipe our own asses. The new generation is falling in line.
I'm in the city and see how kids and many adults left and right of me in coffee shops use chatGPT... many similar stories
I see kids generating cover letters with chatGPT on one side, and on the other is HR using chatGPT to filter through them... how effecient and convenient for all
Last edited:
Elektrishun
Ultra Member
Some while get consumed with AI and I think others will turn away from it. Depending on how it all progresses and what kind of world a person lives in it's probably not a bad idea to have a grasp of what it can do. If I ever find a way to make it work for me, like any other tool, I will probably use it.
One of the office lady's kids girlfriend was fascinated with my truck. She asked if I still used maps, and was blown away when I said all the time. She was even more amazed when I said to get to Calgary from southern Ontario it's basically one road, you "should" be able to find Calgary without a map.
She's a smart kid, but I should tell her how to drive using just a compass sometime 😀