4 ms·

>The cynic in me says that there will be Chat-GPT generation of kids that won't be able to think rationally on their own It’s interesting to me that your impli

by RC_ITR 4y ago

>The cynic in me says that there will be Chat-GPT generation of kids that won't be able to think rationally on their own It’s interesting to me that your implication is “ChatGPT” will give reasonable but wrong answers to things, therefore people will accept those answers. It reminds me quite a bit of “no one will know math because calculators.” There’s plenty of feedback loops that exist for exactly this problem. Kids submit GPT homework and it’s either correct (rendering the skill the homework was testing worthless), or it’s not, and those kids will be punished with bad grades. If anything it will teach a whole new generation how to analyze semi-unreliable text, in the same way Google taught millennials how to search though disparate sources and synthesize answers. >the same way that kids and young people these days are completely unable to have a normal sleep schedule. Conclusions: A lack of empirical evidence for sleep recommendations was universally acknowledged. Inadequate sleep was seen as a consequence of "modern life," associated with technologies of the time. No matter how much sleep children are getting, it has always been assumed that they need more. https://pubmed.ncbi.nlm.nih.gov/22331340/ https://pubmed.ncbi.nlm.nih.gov/22331340/

asimpleusecase 4y ago
On the kids and calculators… in the past week I have had two times that it was clear the person at the till could not make change. In one case the total was £4.90 I gave them a £10 note. Quickly I added a 20p coin so they could give me £5.10 back and save them having to give me so many coins. The person was in the 20’s and I could see immediately they did not know what to do. The phone came out of the pocket and calculator function stated and some tapping proceeded but they did not look comforted and then there was tapping on the till and then I was handed £5 and they rushed to the back room. I would have chalked it up a one off but it happened again a few days later. Admittedly it dealt with an antiquated form of money. However, I would guess that level of innumeracy has to impact them in many way in normal life.
- snozolli 4y ago
  In one case the total was £4.90 I gave them a £10 note. Quickly I added a 20p coin so they could give me £5.10 10 - 4.90 = 5.10. Why was another 20 involved? Edit: I'm guessing the total was 5.10, so you gave 10.20 to get 5.10 back.
- RC_ITR 4y ago
  If your argument is that making change is a high value skill, then we disagree. It's fine that cashiers cant make change when these [0] exist. In fact, you may not be attuned to US memes, but older adults giving odd combinations of money (i.e. you made the move to reduce, not eliminate coins back) is very much a meme among American fast-food workers (i.e. The bill was $4.60 and he gave me $5.20 because I guess he hates Nickles?) [0]https://trinityresources-us.com/products/telequip-t-flex-coin-dispenser/ https://trinityresources-us.com/products/telequip-t-flex-coi...
- pnut 4y ago
  I don't get it either. For a bill of £4.90, you were already owed £5.10 after giving the tenner... How does throwing another 20p into the mix simplify anything? Now they owe you £5.30
danans 4y ago
> Kids submit GPT homework and it’s either correct (rendering the skill the homework was testing worthless) LLMs are trained entirely on content produced by people. If people stop developing the skills required to produce well thought out content, the models will stagnate and even decline. They are completely dependent on humans for content input, so the skill of creating it will always be worth something if the language model is worth something.
- RC_ITR 4y ago
  We agree. Creativity will continue to be more valued, while formulaic synthesis will lose value (in the same way recall has declined in value as databases get better).

by RC_ITR 4y ago

asimpleusecase 4y ago

On the kids and calculators… in the past week I have had two times that it was clear the person at the till could not make change. In one case the total was £4.90 I gave them a £10 note. Quickly I added a 20p coin so they could give me £5.10 back and save them having to give me so many coins. The person was in the 20’s and I could see immediately they did not know what to do. The phone came out of the pocket and calculator function stated and some tapping proceeded but they did not look comforted and then there was tapping on the till and then I was handed £5 and they rushed to the back room. I would have chalked it up a one off but it happened again a few days later. Admittedly it dealt with an antiquated form of money. However, I would guess that level of innumeracy has to impact them in many way in normal life.

snozolli 4y ago
In one case the total was £4.90 I gave them a £10 note. Quickly I added a 20p coin so they could give me £5.10 10 - 4.90 = 5.10. Why was another 20 involved? Edit: I'm guessing the total was 5.10, so you gave 10.20 to get 5.10 back.
RC_ITR 4y ago
If your argument is that making change is a high value skill, then we disagree. It's fine that cashiers cant make change when these [0] exist. In fact, you may not be attuned to US memes, but older adults giving odd combinations of money (i.e. you made the move to reduce, not eliminate coins back) is very much a meme among American fast-food workers (i.e. The bill was $4.60 and he gave me $5.20 because I guess he hates Nickles?) [0]https://trinityresources-us.com/products/telequip-t-flex-coin-dispenser/ https://trinityresources-us.com/products/telequip-t-flex-coi...
pnut 4y ago
I don't get it either. For a bill of £4.90, you were already owed £5.10 after giving the tenner... How does throwing another 20p into the mix simplify anything? Now they owe you £5.30

danans 4y ago

> Kids submit GPT homework and it’s either correct (rendering the skill the homework was testing worthless) LLMs are trained entirely on content produced by people. If people stop developing the skills required to produce well thought out content, the models will stagnate and even decline. They are completely dependent on humans for content input, so the skill of creating it will always be worth something if the language model is worth something.

RC_ITR 4y ago
We agree. Creativity will continue to be more valued, while formulaic synthesis will lose value (in the same way recall has declined in value as databases get better).