smart people doing stupid things

John Larkin snipped-for-privacy@highNONOlandtechnology.com Wrote in message:r

My current project requires iec62304 and it is amusing .

Cheers

Yikes. What does it cost to buy the standard? Does it reference other standards?

Not sure how he managed to say master debaters that many times while seemingly keeping a straight face but it reminds me of this:

It's 345 Swiss franks (USD 380). Probably cites many things, so you may need a bunch of these expensive standards.

It documents the now obsolete waterfall model of software development, at great length, for medical devices.

.

I've had to follow this approach (but not this standard), and it didn't go well, because it didn't deal with practical constraints at all. The electronic-design parallel would be a process that requires that a transistor with very specific properties exist and be available. But in the real world, we have to use the transistors that are available, even if they are not perfect - make what you want from what you can get.

The solution was to design from the middle out, and when it all settled down, document as if it were developed from the top down.

Joe Gwinn

I expect that there will be AI Gurus and with essentially religious cult followers. Another Jonestown is possible.

That's the Microsoft Project Effect: the more tasks you define in a project, the longer it takes.

John Larkin snipped-for-privacy@highNONOlandtechnology.com Wrote in message:r

Only $348, surprisingly it does not reference other standards. At least I dont see any. I got a big 4" binder of paper work that should be sufficient to prove we followed the standard. The problem is getting the old guys to get on board , none of them are interested.

Cheers

Can this be blamed on MS alone? Doesn't "more tasks" mean "larger scope"? My plumber certainly thinks so.

Joe Gwinn

Big process effort. The only thing I know of that is worse is DO-178, the process for development of avionics software that is safety-critical in the sense that failure leads to loss of airplane and all aboard.

I hope you are able remain sane.

Yeah. I'm with the old guys on this. We paid our debt to process and was paroled for good behavior decades ago, and don't want to repeat the experience.

Reminds me of Structured Programming (which forbids Go-To statements): .

Problem was that the Process Police tried to force me to follow this in operating-system kernels. Well, I'd like to see somebody build a kernel without go-to statements.

The deeper problem is that structured programming basically requires that the flow chart can be drawn on a 2D surface without any cross-overs - the nesting must be perfect. Well, good luck following that with real computer hardware, never mind the special hardware that the computer controlled.

Think parallel Finite State Machines interacting and interweaving at random, driven by random external events. Not even a 3D flow diagram suffices.

So in this case, I didn't even attempt to document according to Structured Programming, instead telling the Process Police to buzz off. I only had to show them a real kernel listing once - a wall of assembly code. They had seen only toy examples in textbooks.

Joe Gwinn

Strange that you could know what I was imagining.

Have a look at this and then tell me where you think AI/AGI will be in say

...

You haven't met some of the managers I've worked for.

...

I agree.

I know/have known plenty of people who can do that.

...

What is a decision?

So a decision is a decision. Shouldn't a decision be that which causes a specific fork to be chosen? In other words the current state of a system leads it to produce a specific future state?

I don't claim to know what a decision is but I think it's interesting that it seems to be one of those questions everyone knows the answer to until they're asked.

But then the decision has already been made so why bother with such an examination?

Certainly AI decisions will depend on their training, just as human decisions do. And you can still decide whether to be bound by that decision. Unless, of course, the AI has got itself into a position where it will see you do it anyway by persuasion, coercion, or force. Just like humans do. Human treatment of other animals tends not to be of the best, except in a minority of cases. How do we know that AI will treat us in a way we consider to be reasonable? Human managers often don't. Sure you can make a decision to leave that job but it's not an option for many people.

Actors had better watch out if this page is anything to go by:

In that case I'm not seeing anything different between decisions, goals and choices made by a human brain and those made by an AI system. But what started this was "People are invariably mislead by thinking that there is "intelligence" involved in the technology".

So perhaps I should be asking what is intelligence? And can a computer have it? Was the computer which created these videos intelligent?

But it is still the case that the decision has already been made.

The quality of parenting varies a lot.

In that case how did AlphaZero manage to teach itself to play chess by playing against itself?

I'm not sure I get why it's so essential to have AI explain its reasons. If I need some plumbing done I don't expect the plumber to give detailed reasons why a specific type of pipe was chosen. I just want it done. If I want to play chess with a computer I don't expect it to give detailed reasons why it made each move. I just expect it to win if it's set to much above beginner level. A human chess player may be able to give detailed reasons for making a specific move but would not usually be aske to do this.

Same with humans.

Humans have the same issue. You can't see into another person's brain to see what bias they may have.

I don't see how that's relevant to AI which I think is just as capable of bias as humans are.

Who knows, but whether it's human or AI it will have it's own personality and its own biases. That's why I started this with "One thing which bothers me about AI is that if it's like us but way more intelligent than us then..."

Which is why I started this with "One thing which bothers me about AI is that if it's like us but way more intelligent than us then..."

I think AI will learn the difference between a good or not so good episode just like humans do. Particularly if it gets plenty of feedback from humans about whether or not they liked the episode it produced. It might then play itself a few million created episodes to refine its ability to judge good ones.

Sure.

...

That's the point. I don't. I have to accept a doctor's decision on my treatment because I am not medically trained.

I would have said two million gallons just for the pleasure of watching you go red in the face.

I would have likely given him the job. Those who are good at painting houses aren't necessarily good at estimating exactly how much paint they will need. They just buy more paint as needed.

It's true that you may want to understand their reasoning but it's likely that you might have to accept that you can't.

Sure I do, but I'm very slow to get better at chess. I tend to make rash decisions when playing chess.

I usually spot my error immediately when the computer makes me look stupid.

I don't agree. Learning something like that does not depend on the plumber explaining his decisions.

Sure but so what. Most chess games between humans are not about teaching.

Humans generally don't care much for people they have no personal knowledge of.

If they are, say, a doctor then yes you can ask questions about your treatment but you can't otherwise observe their behavior.

I'm not following what that has to do with AI.

I may not want it but it's possible it could exist. It might be the case that I could do nothing about it.

Oh. Now I get why we're having this discussion.

I don't enjoy having others do my thinking for me but I'm happy to let them do so in areas where I have no expertise.

Who says we have the ultimate authority to ignore AI if it gets cleverer that us?

I would expect it to generate humor the same way humans do.

Same way it might appreciate this:

Yes

I'm not sure I follow that.

Um well they don't sound similar to me but maybe I have a different accent.

That reminds me of a religious teacher many decades ago when we had to have one hour of "religious education" per week for some reason. Typical of his quesions were "why does a calculator never get a sum wrong?" and "can a computer make decisions?". Also typical were statements such as "a dog can't tell the difference between right and wrong. Only humans can." Being very shy at the time I just sat there thinking "there's wishful thinking for you".

So is it ok if I take a step back here and ask whether you think that AI/AGI has some inherent limitation which means it will never match human intelligence? Or do you think that AI/AGI will, at some future time, match human intelligence?

I don't mean to suggest that AI will become human, or will need to become human. It will more likely have its own agenda.