[Elecraft] OT: A.I. couldn't identify a contest

[Al Lorona]

Hi, Dave,

1. Apparently.

2. That's a problem. I already stated that in my original complaint. I suppose a good rule of thumb is: A.I. is not necessarily better than Googling something yourself. You still have to vet all possible answers. That's a rule that not all are aware of. To hear the pundits talking about A.I., you would think that it's going to replace God Himself.

3. Regarding if A.I. could have known about the CQ BW contest... Think about it: I wasn't asking A.I. to go listen on 40 meters to find out what contest it was. I thought it would be able to search all of the contesting calendars out there (which many folks graciously pointed me to) to determine the answer. I mean, several hams here did just that. It wasn't at all unreasonable to expect a decent A.I. algorithm to do the same. If the information is out there somewhere, A.I. should find it. Otherwise, it's useless. As someone here pointed out, there's no actual thinking going on... it's just a really fast algorithm that excels at finding pertinent information related to what you're asking for. In my particular instance, it utterly failed at that.

4. Dave, anecdotal evidence is all I have to go on; it's my personal experience that I'm relating here and basing my opinion on. What other type of evidence could I possibly have? I couldn't possibly know all the prompts that the billions of people living right now have submitted to A.I. and what its track record of accuracy is.

5. That is still to be proved. So far... not ready for prime time.

Al  W6LX

__________________________ 

Dave Gilbert, AB7E, wrote:

1.  Not all AI are created equal.  Some are better at some things than 
others.  Copilot isn't one of the best at anything.

2.  Some AI don't know how to tell you they don't know something. They 
seem to be programmed to tell you something even if it is garbage.

3.  I'm not the dumbest rock in the box and I didn't know what CQ BW was 
either.  Expecting any AI to be accurate on everything is foolish from 
the get go.  Not a single human is either.

4.  I find it humorous that people have such a hangup about AI that they 
jump on any anecdotal example to prove that "A.I. is not even close to 
being ready for prime time" while ignoring how good AI already is at 
some things. I use the paid version of ChatGPT and it is flat out 
amazing in certain situations ... even to the point of anticipating my 
need for things I haven't asked it yet. I find it to be worth every 
penny I pay for it. 5. AI is improving literally month by month. It is 
getting better (not necessarily smarter) faster than any of us are. 
Deriding it out of hand is akin to whistling past the graveyard. Dave AB7E