This source is a data analysis that looked at trump's tweets to see if he wrote the Android ones and his staff wrote the iPhone ones. The Android tweets were angrier and more negative. The iPhone tweets had more links and campaign posts. I thought this connected well to the chapter's point about personas. People assume everything on an account comes from the person whose name is on it, but a lot of the time it's a team. It makes me wonder if people actually care who writes it.

Moving between Dubai and the US, I learned to code-switch before I even knew what it was. The way I talked at school in Dubai, the way I talked with my family, and the way I talked with my friends were all different, but none of them felt fake. That's why I agree with the chapter: code-switching can be an authentic way of expressing yourself. Social media is where this gets complicated for me. Right now I am recruiting for consulting internships, so my LinkedIn is polished and professional, while my Instagram is way more casual and mostly friends. Both are me, but I know a recruiter could easily find my Instagram, and someone from home could see my LinkedIn posts and think I sound like a completely different person. To answer the reflection question, I think most of my changes are authentic, but the context collapse part is where my authenticity gets compromised a little. When I am posting with every possible audience in mind, I am not really expressing myself to anyone, just avoiding saying the wrong thing. My question is whether platforms should be designed to separate audiences better.

Growing up in Dubai, I dealt with this a lot. The UAE doesn’t really use ZIP codes. My family used to put in a random number, and I think that’s a good example of people having to fake their info to fit the system. The form worked, but the data wasn’t actually true.

I think the most interesting part of Agrawal’s thread is how he explains where the “less than 5%” number came from. Every quarter, Twitter took a random sample of thousands of accounts and had people check them by hand. He also said nobody outside the company could recreate the number because the reviews used private info like IP addresses and phone numbers that only Twitter had access to. I think that connects to this chapter really well.

Bay argues that much of the online fight over The Last Jedi wasn’t really fans disagreeing about a movie, but organized political influence efforts meant to make American society look more divided than it is. That shows how bots and trolls don’t need to be the majority.

I get why the book draws this line, but I’m not sure it matters much from the user’s side. . If our ethics questions are about deception and responsibility, the click farm seems closer to a bot than to a normal user. I’d argue the definition should focus on whether the account acts authentically, not on whether there’s code involved.

I don’t think this is true for Kant. He argued that the only thing good without qualification is a “good will,” and that an action only has moral worth if it’s done out of duty, not just because it happens to match the rule.

I think part of why these people had no answers is incentives. Tech workers get rewarded for shipping fast. I think the person building it is the one best positioned to see the risks, so they should carry at least some of that responsibility.