“uhhh actually they can” nice try i just asked it to do 0.1…

don’t be ridiculous computers can’t do addition. they can only NAND, XOR, NOT, NOR etc. it’s ridiculous to imply you could build higher level mechanisms on top of these foundations, its completely implausible

Quoted post hidden · Only visible to signed-in users

3 4 75

Samuel

@samuel.fm

“uhhh actually they can” nice try i just asked it to do 0.1 + 0.3 and guess what. it got it wrong.

0.1 + 0.2, whoops

Not even Python gets this right, sometimes. And LLMs usually get trained with using Python. Now it is obvious why they hallucinate and lie to us! -> docs.python.org/3/tutorial/f...

15. Floating-Point Arithmetic: Issues and Limitations

Floating-point numbers are represented in computer hardware as base 2 (binary) fractions. For example, the decimal fraction 0.625 has value 6/10 + 2/100 + 5/1000, and in the same way the binary fra...

docs.python.org

fry69 3 Oct

I mean: >>> format(0.1, '.17f') '0.10000000000000001' Isn't it obvious that everything goes kaboom when you use such broken languages to train our "super intelligence"? And don't get me started on actual floating point arithmetic in CPUs and GPUs.

Gaëtan Chartrer 3 Oct

Even good old Casio can be more precise