6 Matching Annotations
  1. Last 7 days
    1. The 1980s and 1990s also saw an emergence of more instant forms of communication with chat applications. Internet Relay Chat (IRC) lets people create “rooms” for different topics, and people could join those rooms and participate in real-time text conversations with the others in the room.

      Yeah, IRC was basically the blueprint for a lot of what we still use now — it’s like an early version of Discord/Slack with topic-based “rooms” and live chat. I also think it’s cool how it shifted internet communication from slower, one-to-one stuff (like email) to real-time group conversations where communities could form and evolve on the fly.

    2. One of the early ways of social communication across the internet was with Email, which originated in the 1960s and 1970s. These allowed people to send messages to each other, and look up if any new messages had been sent to them.

      Totally — email was basically the OG “DM.” It’s kind of wild that something created back in the 1960s/70s is still a core way we communicate today, just with nicer interfaces and way more spam.

  2. Jan 2026
    1. In our example tweet we can see several places where data could be saved in lists:

      This section makes it click that a “tweet” isn’t just one thing—it’s basically a bundle of lists (a list of images, a list of likes, a list of replies, etc.). Thinking of it that way also helps explain why social media data gets huge fast, because each post can point to multiple growing lists. It’s kind of wild that even something simple like “who liked this” is literally stored as a list of accounts behind the scenes.

    1. Metadata is information about some data. So we often think about a dataset as consisting of the main pieces of data (whatever those are in a specific situation), and whatever other information we have about that data (metadata). For example: If we think of a tweet’s contents (text and photos) as the main data of a tweet, then additional information such as the user, time, and responses would be considered metadata. If we download information about a set of tweets (text, user, time, etc.) to analyze later, we might consider that set of information as the main data, and our metadata might be information about our download process, such as when we collected the tweet information, which search term we used to find it, etc. Now that we’ve looked some at the data in a tweet, let’s look next at how different pieces of this information are saved.

      This part made me realize how “metadata” can be just as revealing as the main content, even if it seems harmless—like time, location, or who replied. In social media, you can sometimes learn more from patterns in metadata (posting frequency, networks, timing) than from what someone actually said. It also feels like a big privacy issue, because people might not realize they’re “sharing” all that extra info just by using a platform.

    1. We also would like to point out that there are fake bots as well, that is real people pretending their work is the result of a Bot. For example, TikTok user Curt Skelton posted a video claiming that he was actually an AI-generated / deepfake character:

      The “fake bot” idea is wild because it flips the usual problem—now humans can pretend to be AI to get attention, seem mysterious, or dodge accountability (“it wasn’t me, it was the bot”). That makes trust even harder, since it blurs what’s real automation versus just performance. It also makes me think platforms might need clearer disclosure norms, because otherwise people get rewarded for deception either way.

    2. As one example, in 2016, Rian Johnson, who was in the middle of directing Star Wars: The Last Jedi, got bombarded by tweets that all originated in Russia (likely making at least some use of bots).

      It’s kinda scary how coordinated bot/troll activity can make a backlash look way bigger (and more “real”) than it actually is, which can totally warp what people think the general public believes. The stat about a lot of negative tweets being politically motivated or not even human really shows why “online outrage” isn’t always a reliable measure of real opinion. It also feels unfair to creators, because they can get pressured or harassed by something that’s basically manufactured.