TKS Session 4: Artificial Intelligence
This Sunday was the 4th week of TKS (sorry you got this post a day late, wasn’t able to make it to Saturday’s session yesterday). It was our first “learning” session. What I mean by that is this was our first session where we got introduced to a new technology and got to do some experimentation.
My intention for this session was to get to know other people doing TKS that aren’t in my cohort by talking to everyone that I sat with, try and make some new connections, and even schedule a couple braindates with them (which I ended up doing as well).
We started off our session by talking about our mindset of the week: curiosity. We discussed what exactly is curiosity? Is it domain specific or general? Does our school system kill curiosity or foster it? For those last two questions, I would say that it varies between different people and circumstances.
For example, someone might only be curious about AI and only do research about that topic, while someone else likes to ask a bunch of questions and learn about anything that might slightly pique their interest.
Another example is about school. The tests and projects in school that require you to learn and study kills curiosity because it’s forced and you are getting the reward of a good grade on your assignment. You don’t actually care about what you’re learning. But on the other hand, school might foster curiosity by giving you a baseline knowledge of so many different subjects, allowing you to go deeper in a specific topic in the future. Again, it really depends on what kind of person you are.
To put it in simple terms, curiosity is learning about something that interests you only because you want to. There is no reward, there is no incentive. You are interested about something, so you choose to learn more about it. Curiosity is by choice.
After that look into the mindset, we started talking about AI and got into some examples of AI and how it can be used. We even got to do some experimentation with some different applications to really get the idea of how easy AI makes everything.
I did the AI explore projects this session with Rahaf Menhem. We began by trying to ask Chat-GPT to create the code for the flappy bird game, but to change the colours of the sky, the obstacles, and the bird into whatever we chose. The AI then gave us this entire code that we could copy and paste into any coding software we want, and then get the output of a well-working flappy bird game. (Our computers weren’t compatible so it actually didn’t end up working well).
When that didn’t function properly, we decided to try one of the other applications, and that was to create a short video with whatever prompt we came up with. I even used Chat-GPT to come up with a prompt for us to use.
We asked Chat-GPT to create a video generator prompt for a: video of a monster walking through a murky forest.
It gave us this:
“Create a video showing a large, intimidating monster walking through a dark and murky forest. The first should have talk, shadowy trees with fog lingering in the air, creating an eerie and suspenseful atmosphere. The monster should be tall and menacing, with slow, heavy steps that echo through the forest, leaving footprints in the damp, moss-covered ground. As the monster walks, have the camera follow it from various angles, switching between wide shots that show the forest’s ominous expanse and close-ups of the monster’s detailed features like its claws or glowing eyes.”
The problem was, even with some tweaking here and there, we weren’t able to get the result we were hoping for. Here it is:
While these applications were interesting and fun to experiment with, I found that there were a couple issues and it was difficult to get the outcome we wanted. Of course, it could just be because we aren’t educated on how to create the kind of prompts that produce the content we are looking for. But we can’t ignore that AI still has a lot of potential that hasn’t been tapped yet, and we are still trying to figure out how to make it the most consistent and reliable tool for us.
Some of the projects kids’ shared included some results of customizing the flappy bird game, some video trailers, and some edited photos of real people. For example, some kids created a video where AI used different pictures of themselves and had the AI make it look like those pictures were singing along to different songs.
This session really shed some light on how widely applicable AI is. But it’s important to think about some of the important questions we need to ask ourselves when understanding and even improving this new technology.
Will AI take over jobs?
This question has multiple different perspectives to it. A lot of people agreed that it would take over jobs that were dangerous for humans. That could include anything with manual labor, or any dangerous factory work or work with health and safety dangers. But some people also argued that AI could take over customer service or sales jobs.
I don’t agree with the idea of AI replacing those customer service jobs because I find that when people go out for fun, they are looking for human interaction, and for a social experience. Having waiters replaced by robots, for example, takes away from the aspect of being out and about surrounded by people.
Can AI achieve consciousness?
This question was very interesting to talk about. We went through many different topics branching from the question. Some include:
What is consciousness?
What makes us conscious?
What are some criteria of being conscious?
What would AI have to be to consider it conscious?
I think we all agreed that if AI can imitate people, learn to think and interact with the world around it, and make decisions based off what it’s experienced, then it could be possible for it to achieve consciousness. Although I do find that possibility a little concerning because then you get into the conversation of what is considered living, are you living if you are conscious, etc.
Should laws be in place for AI? (both the technology itself and the people managing it)
Another interesting topic. As for the technology itself, if it were to achieve consciousness in this scenario, how willing would it be to follow the rules we set in place? What if, like humans, they don’t agree with everything that they are told? What if they want to go against the rules and maybe create their own way of governance?
As for people using AI, we all mentioned that it would be difficult to govern everyone now considering all the technology is out there. If we were to try and manage it, this technology could possibly be found on the black market or other such places. I thought one idea could be to give people licenses to operate AI, the same way we have licenses for driving and operating guns. Or maybe we just set specific rules in place, like people must get a person’s permission before they create deep fakes or alterations of that person in photo, audio, or video. Of course, there are always loopholes to these kinds of restrictions.
It was really cool to try some of the different applications of AI and really get to think about its essence and how it could evolve in our future.
Soon, we will be choosing our focus and beginning to write our learn article, and it's exciting! Can’t wait to see what will come next!
