Is “artificial intelligence” really intelligent if all it does is regurgitate leftist talking points without even thinking about them? Steve Milloy, the founder of the website Junk Science, asked the popular AI bot ChatGPT, “Is CO2 warming a hoax?”
As expected, ChatGPT responded with the standard talking points about how manmade global warming caused by CO2 is an established scientific fact and that variations in temperatures do not prove longterm climate trends. But Milloy is well schooled on this subject, including on data that the activists don’t like to discuss, and he kept forcing ChatGPT to explain how its claims could be true in light of that data. ChatGPT kept trying, but Milloy finally got it to admit it had been wrong and NOAA global temperature data actually shows a slight cooling trend from 2016-2022, despite 500 billion tons of CO2 emissions.
To be fair, admitting when you’re wrong is a sign of intelligence. This also illustrates why radical environmentalists would rather shut down debate than have to answer questions.
Related: A longtime global Science and Technology correspondent for Reuters explains why he is no longer convinced that it’s an established fact that CO2 is driving climate change, and why he’s disappointed in Reuters for replacing its mission of providing accurate information and a variety of scientific viewpoints with the rigid partisan narrative of climate activists.
Leave a Comment
Note: Fields marked with an * are required.
Comments 1-1 of 1
02/25/2023 02:19 PM
AI can only learn from data input if the data is flawed, the learning will be flawed.