Project Pitches ​

Overall, I really think everyone's pitches were super cool. That being said, I have a few comments.

Assist ​

I'm not sure how this is different from a group chat in the sense that the scope is exactly the same. If I were to ask for help in a group chat, it's the same people who would be responding as in the Assist app, it's just that they would have to check a separate app instead of just getting a ping in the group chat that everyone already uses. In fact, even less people would see it because you are only sending the request to certain people within the group, not everyone.

Second, I have some reservations about the tagging system. Let's say I want to borrow a spatula from a friend. If I tag my request "spatula", it's unlikely that anyone will receive the help request because who would ever think to put in their list of things they are available to help with that they have spatula? If you keep it to generic tags like "lending," that's likewise not very useful because of how generic it is.

There's an enormous barrier when it comes to getting help called the bystander effect. The first part, diffusion of responsibility where people will feel less responsibility when there are more people might actually be worse because they can't tell who else got the request or who else has accepted it yet. The second part, social influence means that people mimic the others around them. With assist, individual might be more inclined to help on their own (reducing negative group influence), but it also might make them less inclined to help in the opposite situation (reducing positive group influence).

Just some things to consider.

TuneTrainer ​

This is a really fun idea and one that I would absolutely use.

I think it's going to live and die off the actual quality of the AI though. Turning words into syllables can be hard and current NLP tools don't get it 100% correct. There are ways around that (you could technically just smush 2 syllables into 1) but you're effectively putting a bunch of AI tools together and each one is going to introduce a little bit of error. Maybe you can have users be able to edit the songs to smooth out the flaws and share them so that if someone creates a really great song, other people can just use it directly? That would add a lot of complexity though, maybe too much for this project.

Also, are there other ways to measure success - potentially cheaper / less involved metrics?

BlueHaven ​

I really like this idea overall. I would love something like this if it was active. The biggest sticking point I have is that I think most of the time students are not going to be that interested in contributing. 6.1040 is pretty unusual in how contributions are part of the grade and depending on some sort of classroom requirement is going to make it difficult to popularize.

I think the app also creates some extra work for the instructors because they have to basically reorganize and repost their content to suit the concept-based format of your app, and also read over / watch any additional resources that students post in order to vet the sources. This is more work than just answering questions on Piazza.

I might be wrong though. I hope I'm wrong and that students will be overjoyed to contribute. I'm just very cynical.