One question I had in this section is whether metadata can affect people more strongly than the post itself. If someone sees a lot of likes or a blue checkmark first, will they trust the post more before they even read it carefully? This made me wonder how much our opinions on social media are shaped by the content, and how much they are shaped by the extra information around it.

I read this article about how numbers may seem clear and objective, even though they do not always reflect the real things they are supposed to measure. What stood out to me most was the part about sentiment ratings, IQ, and aggression scales. The article shows that even something as simple as 2+2=4 can become more complicated when the context changes. This made me think about social media too, because numbers like likes, views, and ratings often seem trustworthy even though they only show part of the picture.

I read C2, which explains how click farms work and how they create fake popularity online. What stood out to me most was that this system relies on real people, not just bots or software, to keep liking and following content across many devices. That connected to the chapter’s idea of “human computers” and made me see false engagement differently. It is not just about technology. It's also about the people doing this repetitive work behind the scenes. I found it a little disturbing to realize how easily likes, follows, and views can be manipulated, since many users, including me, often treat those numbers as signs of trust or popularity.

The example of Tay stood out to me because it showed how quickly a bot can learn harmful behavior online. I don't think this happened only because of racist users. It also seems like the bot was created without enough protection or clear limits on what it could learn. What shocked me most was that it only took one day for Tay to start repeating harmful ideas back to people. That makes this example feel less like a random mistake and more like a warning about what can happen when AI is released without enough care. This made me think that AI can easily reflect the worst parts of people when there are no strong limits.

I don't think it is enough for creators to say that misuse isn't their fault. Even if developers can't control every use of a technology, they still make choices that can make harm more likely or less likely. I also think it matters that laws and society often can't keep up with new technology. When that happens, the people making it have even more responsibility to stop and think about possible harm before releasing it. Deepfakes are a good example, because a tool that can create realistic fake images or videos can easily be used for harassment, spreading false information, or violating someone’s privacy.

This section made me wonder what Confucian ethics would look like in online communities where people don't know each other personally. Can ideas like respect, sincerity, and social harmony still work in digital spaces that are often anonymous and fast-moving?