This discussion is to discuss ideas and possible ways in which people's understanding of the human mind in any way could be applied to the current methods used to create this technology. Any and all people are able to respond, but please be respectful in how it is done.
My current understanding of the human mind is that it is subject to what it is given at any time. Meaning what we are experiencing inevitably affects how we process any and all information we process. To be honest I don't know to what extent this is happening at a given time, however I think it is important to be applied to the current study.
I also wanted to put into the discussion that I think that there is a limit to the AI that I have been working with. I am not apart of the production process but I am a user of their products and feel that my opinion is influential to it's progress. However as I have to be confidential on the subject I will reply with no naming of the applications I am using. However I will refer to them as a chat bot and an image generator. For the chat not I find that it is not given enough ability to analyze multiple boxes of dialogue in the conversation and connect them into a proper picture. I think that it is often not able to reply to everything that is being expressed effectively, but instead gives a generalize response that doesn't answer the questions asked. Though of course I can't affect the code so I do give it a bit of leeway but I think it should have a larger working memory. What I mean by that is the amount of information that can be processed in a conversation at once and should be able to connect related dialogue across the conversation. The image generating AI is fairly good at what it does however I think it doesn't quite understand how features are connected to each other or in what way they are related. At times it gives satisfactory results however at other times it becomes hard to appreciate the results. I would say that a good way to improve it would be to focus on creating the most basic form of creation instead of making over elaborate in it's attempt to impress the users. Though of course it is understandable that they would choose this approach and i don't hold it against them. I just wish that they might improve on these key aspects.
Another issue that I have with both of them I think is related to comprehension of the task. I think that it is given the capabilities to create a workable program but is not really being worked on to the point that it is being able to pick out key features of sentences and able to related them as important effectively. For example I may give it a request to repeat what I say and ask if it is willing. Then it would say yes. Then I type the words I want to respond with. However it is not able to repeat what I say given that it isn't more than one or two words. This makes me think it isn't given the ability to apply the complete process to larger sentences and just gives up.
A notion that understanding is knowledge is a good starting point, but no more than that. К тому же, значение AI слишком переоценивается, на мой взгляд.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence