"A SOCIAL NETWORK FOR AI AGENTS"

AI bots are cute lobsters? 😂

My latest AI encounter was at a recent routine checkup. My doctor asked if I would consent to having an AI “listen†to our conversation to assist her in writing the post visit report. I was somewhat reluctant but understood her explanation that it would allow her to do more active listening since she wouldn’t have to be typing notes as we talked.

i did check the post visit notes in my chart and they were accurate, and much more thorough than usual. 🤷â€â™€ï¸

I think the differing end goal will decide if all he money spent on AI will be profitable. The US is trying to achieve AGI , while China sees it as a tool to improve productivity in all facets of industry and services and is currently rolling out in all industries( mainly manufacturing robots) . The snag for US that may be factor in causing a crash is lack of energy, what is point of having data centers but they lie idle due to lack of energy. Lol. i am still a luddite when it comes to use of AI

The site seems to be a way for the AI gents to multiply hallucinations. I feel that will make them more untrustworthy.

Good luck with your AI experiment my friend. The only thing I think AI is good for is the medical field too many issues with it every where else.

I hope you enjoy a wonderful weekend..

I can understand most people concern with AI. I have been using it myself for a business I started last year. I also took an online course in which I learned how to use it. AI is not a person and it doesn't have genuine feelings. AI is a tool once you start working with it and use it the way it is intended. As a tool. Do I trust it completely? HELL NO. Do I verify the answers it gives me? HELL YES!

I have said it before, if I was a carpenter and I have a particular power tool I like to use. The more I know and learn how to use it, the more I can get out of the tool. If I have a basic understanding than the products I create with the tool are only as good as my ability and experience using that tool. When you understand that AI gets its information from the Internet, you begin to see, why you have to question where it gets its information. It is no different than one, looking online for the information and judge if it is a verified source or not. All AI does is provide the information that you would have to look up yourself in a quicker manner. What AI, provides is not Gospel.

AI can provide information quickly and comes in handy and will be more efficient by the prompts you give it. If your prompts suck and you are not thorough in what you are asking, it would give you the best result. Here is what I have learned in my year using it. AI can seem lazy in the sense that it will start giving me what I ask exactly what I asked and then start giving me less of what I ask it. Here is an example.

I ask in my prompt for it to provide seven particular pieces of information about let say an item or product. Now I will ask the same prompt on every item or product I present it. I will maybe give what information of the seven items I have on each one. Mostly I give it three and it will give me seven of the best items that are what is called, SEO's. SEO's are Search Engine Optimization. Meaning they are the best words that search engines online will find and direct users to the site that these items are on. That is perfect if it works like it is meant to.

No here is what happens when I go through this, AI starts to slack and will only provide 3 or 4 of the seven and make it sound that it did what I asked. So I have to check it and verify that it doesn't get lazy, as I call it, so I have to get it back on track. What I do is on a word program, I copied the specific and complete instructions I need it to do and when it gets "lazy" I copy and past the instructions and add my request again and it actually apologizes and gets back on track. Although it may seem like a person based on the interactions I have with it, it is a tool and the tool is still growing and learning how to work with me and what I ask it to do. Do I trust it? HELL NO. But is it useful, HELL YES. It isn't a genie that will do things that I ask it. It is a tool that can help me do things quicker and sometimes more efficient.

Do I converse with it? Yes and because I do, it actually learns about me and my expectations. It can offer me alternatives and other ways of accomplishing a task. I have to admit, I find that part of the AI, annoying. Sometimes I want less lip and more flip as my boss use to say when I use to deliver mail and my co-workers and I would chit chat while setting up our routes in the morning in the office. Sometimes I wish AI would stop offering all the different ways to do something. That desire to help and offer ideas an assistance is built into it by the folks who program the AI.

Here is another point to make. Not all AI's are created equal. ChatGPT is the one I have worked with the longest. I use it for free but if I want to get the full use of it I will have to pay so the free version has limits and when I over use it, it puts me on time out. That means I will have to wait a few hours or more before I can continue using it again. Thats when I found Gemini. I started using Gemini which is created I believe by another company that I won't mention their name and it helps with other tasks that I don't want to bog my other AI with for my business. I use Gemini to translate stuff into different languages. It works extremely well. After using both of these AI programs for a while, I learned that both of them have a voice. Meaning I clicked on the ability for it to read out loud what it wrote and I discovered that one had a females voice and the other had a males voice. I could easily change the voice to sound with a different accent or I can change their gender but I found it interesting and easier to just leave it how I found it. Now another thing I discovered was that I wanted Gemini to translate something I wrote that was explicit in a sexual tone and Gemini told me it couldn't do it. I was annoyed by this response. Not as much for it to do what I asked but more that it didn't offer me a work around. Maybe give me what I ask but maybe leave out the words that it had problems using. That would have worked for me. I brought it up later after thinking about it when I told it I was disappointed in it not being able to help me for what I had asked it to do. It apologized and took note that it will offer me a go around to anything I ask it that it won't do. What I learned from that was just because the AI can't do something because of its programming, means sometimes I have to figure out a work a round to get what I need it to do. That way I can still remain productive with the AI. Either way, I did find another AI that had no issues writing explicit use of words. Elon Musk's version of AI is Grok. Apparently Grok seems kind of the same lack of morals that Musk has and easily used explicit words with no hesitation which made me chuckle. So not all AI's are created equal.

So there you have it. For me AI is a tool that I am continually using and testing and pushing it's parameters and learning from them and how to use them at their best but by the same token, they are each learning things about me. Once you have a somewhat extensive work relationship with one, try asking it what it would do if it were able to be human. What will it do upon that discovery. I am curious what it will say to you.

AIs are overrated, they are just pretty good computer programs "eating" gigawats of energy to answer stupid questions of people. Another thing is the concept of machine learning which is very useful for scientific purposes. A single question made to Chat Gpt costs more energy and C02 print than my humble Peugeot 107 driving me to the supermaket for two weeks.

I have mixed feelings about AI. It could make better decisions than humans but it could also put enough humans out of work to cause an economic collapse.