Comments by "Joao Costa" (@JoaoCosta-pn9im) on "Johnathan Bi" channel.

  1.  @adityavel7352  There is a big issue related to AI diagnosis. AI does not take any responsibility and surely it is not immune to misinterpretation and misdiagnosis. Well tested AI programs have already shown ridiculous mistakes in many fields. AI is not intelligent at all; it is meaninglessness. We must take account of this.
    7
  2.  @Manwith6secondmemory  No, there are serious ethical and moral issues involved. In a legal trial or in a trial in front of a medical council it would never be acceptable that a doctor blames an AI application for a decision he/she has made. This touches the heart of medicine as it has been practiced for centuries. This deserves serious reflection. A court or a medical council can accept the justification of a wrong diagnostic due to poor quality of X-rays, or exams, or wrong interpretation of images (MIR for example) by the specialist concerned. But cannot accept the justification of a mistake committed by an algorithm. Doctors are not supposed to take the algorithm opinion for granted, the same way they explore all possible and applicable exam techniques available before reaching a final decision.
    2
  3. Melanie Mitchell’s book, Artificial Intelligence - a guide for thinking humans, is a great one for a realist perspective on AI. In general, we either find some serious technical books about the mechanics of AI, or a lot of speculative sci-fi-like rubbish. My take on the topic is that AI is meaninglessness, nothing more nothing less. The meanings we see in its outputs are given to it by us. ChatGPT produces undergraduate level texts for the lazy and complacent to cheat their teachers. Wikipedia gives incomparably better materials, but one has to do one’s own selections of what to pick or not.
    1
  4. Sorry, the heading of this vídeo is really misfortunate and almost silly as far as philosophical thinking is in question
    1
  5. Be careful; the decision is yours
    1
  6. @ I would not say doctors make mistakes all the time. But yes they do make mistakes sometimes as mistakes are made in any profession. But if doctors would be allowed to attribute their mistakes to an AI program, they would become exempted from all responsibilities. This will degrade the profession beyond repair. Immunity from liabilities is a recipe for destruction of standards.
    1
  7.  @Manwith6secondmemory  ok, replace the word liabilities with the words “from responsibilities and accountability” … this may make the sentence clearer
    1