Is that any relation of 'fuzzy logic', aka woolly thinking? *
*Apologies to the late Sir Terry
Printable View
No different to a lot of IT programming which runs the decision making process for so many things in life. Is based on the premise that the past is a good basis for the future which will be the same as the past. Then they wonder why it has trouble when the world does not behave like that. When the difference is not too great those impacted are just collateral damage and the cost can be adsorbed or extracted from the customer impacted. Larger movement away from what went before causes real problems and costs that are big financial losses
I had a play with chat GPT late last year and got it to write a response to an English essay question that youngest daughter had been procrastinating about doing. It was very good at providing a response to the questions , most of which were "compare the human experience in book/movie A with the human experience in book/movie B in XXX words" Chat GPT did an excellent job in writing a couple of paragraphs of convincing words, the irony was that youngest daughter learned more about the aim off the exercise by determining the quality of the chat GPT response than she would have done by writing her own essay.
The current moral panic around chat GPT reminds me of primary school in 1975 when the first calculators became affordable and the education community was certain that no-one would be able to count anymore because of the demise of long division.
Regards,
Tote
Funny that... I find most kids ARE having problems counting :)
I am not worried about AI making schooling worse, it has been on a steady decline over decades so far as I can determine, no I think your example points clearly to the benefits. What I am worried about is the answers that are given are, as some before me have opted to call it "reality" or "consensus", which might not be entirely accurate or complete.
My take on it has always been (education that is) that critical thinking should be far higher up the priorities list for kids so that they can question some bloody commercial trying to tell them stuff etc. As AI becomes seemingly more human, the same goes for AI.
-P
Or when I was in primary school and ballpoint pens first appeared in Australia at the end of the 1940s. They were flat banned, as their inability to give different stroke widths compared to a 'proper' pen would stop us from learning to write properly (we used 'proper' pens and ink, with ink wells in the desk, filled daily by a trusted kid called an "ink monitor".) and they needed to be held more upright.