Google's take on memory safety

Dom DiSc dominikus at scherkl.de
Thu Mar 7 10:40:21 UTC 2024


On Thursday, 7 March 2024 at 08:12:36 UTC, aberba wrote:
> Also depending on the task, the results can be better. For 
> example it appears natural language processing has progressed 
> better than other AI tasks/fields.

That's only because for natural language there is better training 
data, because humans are good at natural language. (Still AI is 
not as good at it as humans, only not so bad anymore that one 
shruggs in disapointment).

But in programming languages humans are not as good, so the 
training data also is not. I have no high hopes in using such an 
AI.


More information about the Digitalmars-d mailing list