Microsoft is building a new AI chatbot that will replace Bing, Teams, and Outlook.

Just like AutoCAD is getting one so they can compete with DALL-E. Tech is getting to be a lot like Star Trek movies, where the “scary bad thing” is AI. But this one makes me nervous. Just like I can AutoCAD, DALL-E can also be used on coded files. But I still have to give DALL-E at least 20 parameters by typing, which means you have to have some basic knowledge of the thing you need to “gimme” before you can ask for AutoCAD to make the thing. You would have to do the same when using a DALL-E coded file. But now both apps will soon have a chat interface, so they can have parameters given to it. But I saw an example of something like this with the chat AI on a site in India. Their AI is like the best proofreading AI I had ever seen. You could give 1,000 characters to the AI. And AutoCORRECT the corrections, so they are all simultaneous, rather than chronological. But AutoCORRECT is a cheap product, compared to the new AutoCAD “add on” that is going to cost $9 per month… Which is also what the DALL-E chat is going to cost. But what is the ethical line, where the AI is smart enough to go with the simultaneous corrections, rather than chronological ones. AutoCORRECT does have something like this with the “explain yourself” option, where you can explain how you want to spell a word to make it safe. But the chat is going to be more human-like. But I have seen with the chat, in India, that the AI is already advanced enough to go straight to simultaneous corrections to make a safe spelling and grammar option, where the human itself knows the “right” way to spell it. Even “dictation corrections” are more advanced than what I see in American AutoCORRECT programs, where they are able to use the same rules for dictation simultaneous corrections, into written spelling corrections corrections, where the human narrator knows the correct spelling, so the AI can learn to gradually do the corrections, as the human is speaking into the dictation software. So what is the ethical line, when the AI is more advanced than a human, and corrections can be done simultaneously without needing to go through chronological corrections, to make sure there is no errors in the rules of corrections? Or, is it the ethical line when the corrections are so advanced that they are able to make changes OUTSIDE of the chronological corrections, where the AI can say, “I know this sentence in the paragraph is saying one thing, but I know what you actually want to say is something else,” so the AI is smart enough to go with the simultaneous corrections OUTSIDE of the chronological corrections. But AutoCORRECT is a cheap cheap program, compared to the new AutoCAD chat add on. But the cheap program is still able to do chronological corrections, if you choose to do corrections, which are only a little bit more advanced than what the basic AutoCORRECT has had for over a decade. So what is the ethical line, where the AI is smart enough to make simultaneous corrections, “better” than humans, but “just enough human-like” not to “lie” and make AutoCORRECT changes outside of the chronological corrections.

Microsoft is building a new AI chatbot that will replace Bing, Teams, and Outlook.

Comments (0) 2 👁️