OPENING QUESTIONS: Let's say you were working very late taking observations and the software that connects the ZWO ASI 1600 mm tracking camera to the Sky Watcher EQ6-R Pro Mount is misbehaving.
You've only got tonight to get your work done and the night is quickly passing by. You google the situation and get a number of links to message boards.
You start reading through them and get lots of conflicting advice.
You're starting to get very frustrated as the night moves on... what do you do?
OBJECTIVE: I will practice problem solving through interactive conversations during today's class.
WORDS FOR TODAY:
WORK O' THE DAY:
You don't remember a time before GOOGLE search came along. The internet pretty much exploded when the first browsers appeared in the early-to-mid '90s.
There were a whole bunch of 'search engines' such as Alta Vista, Lycos, Web Crawler, Yahoo, Ask Jeeves (Really?) and others.
None of those were particularly better than any of the others so users typically got used to using one search engine and then stuck with it.
Results were easily manipulated by web designers and it was a bit of a slog to find any useful information without getting mauled by misdirection, pop-up ads and all sorts of crap.
That's not to say it was unusable since you could still get useful info, it was just kinda frustrating at times.
Google came along with a new search algorithm (method) and blew all of those out of the water.
There's a reason why 20 years later that Google still rules the roost (ever use Bing?) with over 90% of the world's (I think, but certainly the U.S.) search fussers.
We have all been conditioned to use Google since it is so helpful.
However, there is a better way to problem solve online.
AI
Now don't get me wrong, I'm a huge NON-fan of AI to do work. I think it is a very poor substitute for human generated content. It can be a crutch and a lot of people fall back on it and that drives me a bit crackers.
Having said that, AI can be an AWESOME *interactive* tool to solve problems.
It kind of amazes me that we aren't teaching our students how to have interactive conversations with AI.
I first used it last summer in the situation outlined in our opening question today.
The trick is to force the AI NOT to throw massive amounts of text at you but instead to have a back and forth conversation. Treating the AI as a knowledgeable source that you can talk to is most helpful.
To wit:
- Introduce yourself to the AI:
"I'm a high school junior in an astronomy class. We are working on a project about dark matter and I need some help understanding that"
AI:
"Dark matter is a type of matter that doesn't emit, absorb, or reflect light, making it invisible to telescopes. It interacts with regular matter through gravity, and scientists study it by observing its effects on galaxies and galaxy clusters. What specific part of dark matter are you trying to understand?"
"Wait a second, my teacher told us that we don't really know what dark matter is and you are talking like it is well understood."
AI
You're absolutely right—dark matter is not well understood, and scientists don’t yet know what it actually is. The term "dark matter" describes something we observe indirectly through its gravitational effects, like the way it affects galaxy rotation and the bending of light around massive objects (gravitational lensing). Theories about its nature range from being exotic particles to modifications of gravity itself. Would you like to explore these ideas or focus on how scientists detect its presence?"
Remember, you can (and SHOULD) control the 'tempo' of the conversation. chatGPT tends to throw walls of text at you. You can tell it not to do that:
You're throwing kidn of a lot of words at me, can you slow down and go sentence by sentence?
AI:
"Dark matter is something we can't see directly.
Scientists know it exists because of how its gravity affects stars and galaxies.
We don't know exactly what it is, but there are some ideas about it.
Would you like me to explain how scientists find evidence of it?"
Notice that chatGPT has added double spacing and has cut its answer back a bit.
"I still don't understand. Our teacher says we really don't know what it is. How can we not know what it is and still get evidence for it?"
AI:
Good question! Here's how:
We don't see dark matter itself, but we see how it pulls on things with gravity.
For example, galaxies spin faster than they should based on the stars and gas we can see.
The extra gravity holding galaxies together must come from something invisible.
That "something" is what we call dark matter, even though we don’t know what it’s made of.
Notice that GPT is adapting to a much more 'conversational' tone
"That makes sense. I remember when we were studying the solar system, we learned that in the 1920's were sure there something out there past Neptune because of changes in Neptune's orbit."
AI:
"Exactly! That's a great comparison.
Just like astronomers noticed Neptune’s orbit was being influenced by something unseen, scientists now see galaxies behaving as if there's invisible mass pulling on them.
In Neptune’s case, they eventually found Pluto (though it turned out not to explain everything).
For dark matter, we haven’t found a "visible" explanation yet, but the gravitational effects are just as real.
Would you like to explore more examples of how dark matter is detected?"
You have to be ready, willing and able to CHALLENGE the AI when you disagree:
"I'm going to push back on that. My teacher said we haven't detected dark matter."
AI
Your teacher is correct—we haven’t directly detected dark matter itself.
What we’ve detected are its effects, like how galaxies spin or how light bends as it passes near large groups of galaxies.
These effects strongly suggest something massive and invisible is there, but we can’t observe dark matter directly yet.
Does that clear up the distinction?
"Yes, I understand now"
══════════════════════════
Final Note:
If you have an account with GPT it will REMEMBER you and it will remember previous conversations which can be super helpful