Forum Discussion
Bing Chat / Copilot "Invents" fake "facts" - "I don't know" should be a valid answer!
AI developers introduce significant limitations that affect access layers. This makes it possible to automatically verify critical errors. It really works!
AI can be compared to our cars. We all use them, even though they break down and can be dangerous.
This is an interesting blog takes up philosophical topics, please give me your opinions.
The Monkey and the Dolphin – Using Human Intellect in the AI age (microsoft.com)
You are right!
I've noticed that the comparison between a car and an AI is clearer than the monkey-dolphin analogy.
A person has to make a choice (your answer showed that you have to consciously use advanced technologies).
AI can certainly influence our decisions indirectly, prompting us to come up with an easier solution that doesn't require effort.
- Which is worse, I wonder: a very popular car with glitchy AI features a driver relies on to much, or your average distracted driver?
You're talking to a mother who taught two kids to drive and turned off the side and backup cameras during lessons. And the wife who tells her husband she's blaming HIM, not the automatic breaking, of he hits the car in front of us. 😄 You can imagine how unkind is be about drinking or texting and driving. 🤣 Car manufacturers are equipping their vehicles with more and more safety systems that are supposed to make all tasks easier for the driver!
The concept of AI is similar.
Deleted wrote:AI developers introduce significant limitations that affect access layers. This makes it possible to automatically verify critical errors. It really works!
AI can be compared to our cars. We all use them, even though they break down and can be dangerous.
This is an interesting blog takes up philosophical topics, please give me your opinions.
The Monkey and the Dolphin – Using Human Intellect in the AI age (microsoft.com)
A couple of thoughts: How often do you impose significant new limitations on your car in order to maintain control over it, even as they degrade aspects of its performance? Does your car require layer upon layer of different braking systems?
Trust isn't something that can be built with a one-size-fits-all framework, it's an emergent state, largely dependent on other considerations, many of them subjective. Unfortunately, if someone wants to inspire trust in their product, they would probably get more mileage out of a good PR team than a good ethicist.
Artificial intelligence is all the rage right now, but we all know that fashion is changing!
I don't talk to the computer -> I choose to talk
with other people.
- I agree that "artificial intelligence" is a poor choice of names for the tool, and part of the problem is that too many people don't much like to think for themselves and have been conditioned to trust the reliability of computers without having been taught "garbage in, garbage out."
I've done text mining using NLP and have an idea of what it would take to do it really well and an idea of what happens when there's not enough time and money put into it, but people still have an inclination to trust it. They have the "trust" part down, but don't know how - or are unwilling - to "verify." When the AI starts feeding on its own flawed output, I think it will just be an infinite loop of garbage in and out. Maybe it's the AI that's too trusting. Thank you - I guess I'm a monkey 🙂
But for me, the most important thing is to build trust - which is limited by the human intellect ( it's a real security.
AI is just a tool (the name artificial intelligence is not true to me)
- I need coffee and a second read before responding, but it may be too much (and to credit the average human operator otherwise may be disingenuous) to expect them to turn their heads in time.
Have you ever seen a dolphin that COULD turn it's head to see the monkey on it's back?
