2 Antworten auf „"Informelles Beratergremium" des RKI: Auch ChatGPT hat Bedenken“

  1. AI 'spills the beans' about the Covid-19 event
    A chat with Perplexity AI reve­als a shocking lack of 'bias'

    “A fri­end asked me what I thought about the new AI system cal­led Perplexity, after he had read an artic­le tit­led “Finally, the Search Engine Better Than Google“, which dis­cus­ses a pod­cast whe­re Lex Fridman inter­views Aravind Srinivas, CEO of Perplexity.

    As I am sure you are awa­re other AI tech­no­lo­gies, such as ChatGPT, make stuff up, inclu­ding fab­ri­ca­ting refe­ren­ces to fic­ti­tious rese­arch papers. This flaw is well known and is an exam­p­le of whe­re AI suf­fers from so-cal­led: ‘hal­lu­ci­n­a­ti­ons’. Hence unless you alre­a­dy know and can trust the sources used by the AI, any reli­ance on the ans­wers pro­vi­ded by the AI, wit­hout this neces­sa­ry pri­or exper­ti­se, might be sever­ely mis­pla­ced (to put it mildly).

    I won’t offer any detail­ed con­clu­si­ons, except to say that it seems incre­di­bly accu­ra­te, and more important­ly is ‘awa­re’ of the com­pe­ting expl­ana­ti­ons behind the Covid-19 event. It also isn’t shy about fin­ding and using sources that sup­port the­se com­pe­ting expl­ana­ti­ons. However, despi­te occa­sio­nal, and obvious, attempts to deploy ‘defen­si­ve mea­su­res’ against my que­sti­ons, it abject­ly fails.“

    https://​whe​rea​re​then​um​bers​.sub​stack​.com/​p​/​a​i​-​s​p​i​l​l​s​-​t​h​e​-​b​e​a​n​s​-​a​b​o​u​t​-​t​h​e​-​c​o​vid

Schreiben Sie einen Kommentar

Ihre E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert