3 ms·
After swapping an and b in your example it correctly detects the divide-by-zero, impressive. After replacing the zero with an expression that must be mathemati
by DaiPlusPlus 4y ago
After swapping an and b in your example it correctly detects the divide-by-zero, impressive.
After replacing the zero with an expression that must be mathematically evaluated to get zero, it still detects the divide-by-zero - that’s downright spooky.
https://whatdoesthiscodedo.com/g/6a8f359 https://whatdoesthiscodedo.com/g/6a8f359
ChatGPT’s inability to verify its claims limit its usefulness - but the example above is unsettling because it certainly seems to be evaluating things somehow, so what’s stopping OpenAI from enabling it to verify certain things through evaluation?
- agent281 4y agoI'm impressed with how well it catches the divide by zero error. I obfuscated it a bit by using subtraction, bitwise and, and using randint from python, but it still catches the error: subtraction: https://whatdoesthiscodedo.com/g/b12eef2 https://whatdoesthiscodedo.com/g/b12eef2 bitwise and: https://whatdoesthiscodedo.com/g/95a1492 https://whatdoesthiscodedo.com/g/95a1492 randint: https://whatdoesthiscodedo.com/g/8cf88de https://whatdoesthiscodedo.com/g/8cf88de
- ChatGTP 4y agoAre you sure it’s not just assuming and not evaluating, called your bluff and now you’re freaked out ? Sucks you can’t be sure? I had a look at your code, it is really cool it can do that but I’m sure there are tens of thousands of examples like that in it’s training data, the example using rand was also cool but again, I can imagine it’s seen that before a fair few times too?