Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)IM
Posts
2
Comments
3
Joined
1 mo. ago

  • Well yes but also no. Every text will be potentially wrong because authors tend to incorporate their subjectivity in their work. It is only through inter-subjectivity that we can get closer to objectivity. How do we do that ? By making our claims open to scrutiny of others, such as by citing sources, publishing reproducible code and making available the data we gathered on which we base our claims. Then others can understand how we came to the claim and find the empirical and logical errors in our claims and thus formulate very precise criticism. Through this mutual criticism, we, as society, will move ever closer to objectivity. This is true for every text with the goal of formulating knowledge instead of just stating opinions.

    However one can safely say that Chatgpt is designed way worse then Wikipedia, when it comes to creating knowledge. Why ? Because Chatgpt is non-reproducible. Every answer is generated differently. The erroneous claim you read in a field you know nothing about may not appear when a specialist in that field asks the same question. This makes errors far more difficult to catch and thus they "live" for far longer in your mind.

    Secondly, Wikipedia is designed around the principle of open contribution. Every error that is discovered by a specialist, can be directly corrected. Sure it might take more time then you expected until your correction will be published. On the side of Chatgpt however there is no such mechanism what so ever. Read an erroneous claim? Well just suck it up, and live with the ambiguity that it may or may not be spread.

    So if you catch errors in Wikipedia. Go correct them, instead of complaining that there are errors. Duh, we know. But an incredible amount of Wikipedia consists not of erroneous claims but of knowledge open to the entire world and we can be gratefull every day it exists.

    Go read "Popper, Karl Raimund. 1980. „Die Logik der Sozialwissenschaften“. S. 103–23 in Der Positivismusstreit in der deutschen Soziologie, Sammlung Luchterhand. Darmstadt Neuwied: Luchterhand." if you are interested in the topic

    Sorry if this was formulated a little aggressively. I have no personal animosity against you. I just think it is important to stress that while yes, both may have their flaws, Chatgpt and Wikipedia. Wikipedia is non the less way better designed when it comes to spreading knowledge then Chatgpt, precisely because of the way it handles erroneous claims.

  • Kann ich auch nicht recht verstehen. Aber schreib doch mal dem Autor einen freundlichen Leserbrief und frag nach, oder ? Ich bin auch gespannt, ob er einfach nichts davon weiß oder es bewusst auslässt um diejenigen Leser:innen nicht zu verschrecken, die sich als "nicht so technikaffin" verstehen.

  • ObsidianMD @lemmy.world

    Obsidian + Git on Android

    Netzkultur / Netzpolitik @feddit.org

    SZ fordert Fediverse ähnliche Plattform von ARD und ZDF | Digitale Öffentlichkeit - Achtung, sonst sind wir offline