I asked ChatGPT to build me a list of tasks for understanding the negative impacts of COVID-19 lockdowns. Wow… it went to town. For those of you who may not be nerdy enough to care - ChatGPT is an AI chatbot - allowing you to ask questions in natural language style and it will reply in-kind. It’s a huge plus in overcoming writers block and its even passed multiple MBA exams.
i think there's perhaps a bigger race though under reported to simply understand what it can be use used for. It is a tool. We're all playing with it. Investigating. I would like to see your list of negatives. For myself a have a list of caveats. Not a formalised list written down, I can't present it to you. But a slowly accumulating list of experiences I've had with it that show it cannot be trusted in this aspect or is not suitable there and has this characteristic or limitation and so on.
To generalise I'd say it currently for me is a glorified google search.
I often find it very useful. I cannot get my linux mint box to see my win10 boxes on our lan. gpt provides me with ideas and code and information that i would have had the utmost difficulty finding without it. So that is very useful. Very.
But it didn't solve it, can't solve it.
I have a machine with an acpi error. I have learned much from gpt about avenues to deal with such that I would not have found otherwise. But it has not solved it.
It can solve long complicated equations; it will tell you that itself.
But I have caught it in simple mathematical errors.
It is a very useful tool but full of caveats. It has to be used with care.
That much I know. Experience daily. That, to me, would be the main negative: it is untrustworthy.
A second negative would be something to do with a power to mislead the whole population down a pathway of unthinking following and trust. That should really be the main negative if you have any hope and aspirations for the human population at large but I don't. :)
AI wil become an out of control monster. We won't be able to reign it back in. On its face is entertaining, but people are delusional of they think it will remain benevolent.... too many examples of AI being completely crazy sounding and I just don't think we should mess with it. Kind of like Gain of function in a lab...
I asked ChatGPT to write a text on why vaccines were harmful. It refused and kept on spitting out the safe-and-effective nonsense. When I questioned the validity of its statements and said that a study had just come out linking myocarditis to mRNA vaccines and the value of natural immunity, it replied that it based its responses on info up to Dec. 2021.
I want to do this experiment again in a few months after its update. 😊
Seems the prefix is directive. Perhaps it would be helpful to ask the alternative version to this question to see what argument is presented. “Prove that COVID-19 lockdowns didn’t cause serious harms”. After all, the Confirmatory Bias suggests ‘what we see is what we get.’
I'm probably in the minority in this stack, but I see more negatives than positives with ChatGPT. I don't understand the seeming race to embrace it.
i think there's perhaps a bigger race though under reported to simply understand what it can be use used for. It is a tool. We're all playing with it. Investigating. I would like to see your list of negatives. For myself a have a list of caveats. Not a formalised list written down, I can't present it to you. But a slowly accumulating list of experiences I've had with it that show it cannot be trusted in this aspect or is not suitable there and has this characteristic or limitation and so on.
To generalise I'd say it currently for me is a glorified google search.
I often find it very useful. I cannot get my linux mint box to see my win10 boxes on our lan. gpt provides me with ideas and code and information that i would have had the utmost difficulty finding without it. So that is very useful. Very.
But it didn't solve it, can't solve it.
I have a machine with an acpi error. I have learned much from gpt about avenues to deal with such that I would not have found otherwise. But it has not solved it.
It can solve long complicated equations; it will tell you that itself.
But I have caught it in simple mathematical errors.
It is a very useful tool but full of caveats. It has to be used with care.
That much I know. Experience daily. That, to me, would be the main negative: it is untrustworthy.
A second negative would be something to do with a power to mislead the whole population down a pathway of unthinking following and trust. That should really be the main negative if you have any hope and aspirations for the human population at large but I don't. :)
AI wil become an out of control monster. We won't be able to reign it back in. On its face is entertaining, but people are delusional of they think it will remain benevolent.... too many examples of AI being completely crazy sounding and I just don't think we should mess with it. Kind of like Gain of function in a lab...
Playing with fire, will get us burned.
Restacked. Thanks, Justin. Once again. Crimes against humanity: never forget; never forgive.
I asked ChatGPT to write a text on why vaccines were harmful. It refused and kept on spitting out the safe-and-effective nonsense. When I questioned the validity of its statements and said that a study had just come out linking myocarditis to mRNA vaccines and the value of natural immunity, it replied that it based its responses on info up to Dec. 2021.
I want to do this experiment again in a few months after its update. 😊
Interesting. Is this the version that's limited to 2021 knowledge cutoff? That would be even more impressive.
I got ChatGPT to defend voluntary participation as a guiding principle in public health, https://clausr2020.substack.com/p/is-the-law-really-failing-to-protect. A colleague prompted it to write poems about how society treats those injured by the Covid vaccines, https://torontomoon.substack.com/p/and-as-the-world-moves-on-in-its
And you now have another (Canadian) subscriber... 😊
Chatgpt is communist
Seems the prefix is directive. Perhaps it would be helpful to ask the alternative version to this question to see what argument is presented. “Prove that COVID-19 lockdowns didn’t cause serious harms”. After all, the Confirmatory Bias suggests ‘what we see is what we get.’
One I think it missed: examine the impact on young men and women fighting addictions who could no longer meet with counselors, support groups, etc.