I distinctly remember Donkey Kong Country as my first gaming experience. When my parents needed a babysit, they would often let me sleep at my aunts place. My older niece and nephew also lived their, but the age gap was quite big making it not ideal for us to play with toys together. One day however, my nephew had a SNES from his friend there and he was playing Donkey Kong Country on their TV. I remember being completely entranced by it and being unable to put it down (even though it was very difficult for me at the time). From then on I was always hoping that the "Gaming Machine" would be present if I stayed over, which was often the case as they figured out that this was a very easy way to keep me occupied. I later got a green Game Boy Color, and of course this was my first game for the system. I played it countless of hours, and even though I later got a Game Boy Advance SP, this game would remain in my rotation until I got a DS many years later.
I've barely played any of the later installments. I got Donkey Kong Country 2 for the Game Boy Advance when I was young, but found it to difficult and didn't really like the new protagonist as much. After my DS I became a playstation fanboy for the rest of my childhood and teens.
Now that I have bought a 3DS I've started playing Donkey Kong Country Returns. It's really nice, but I found it a bit overwhelming and haven't really touched it since.
Also a shout out to the Game Boy game called Donkey Kong, in which you actually play Mario with some incredibly varied platforming for the time. An all time classic!
I hate AI. Why?
However
I also took the time to read the original blog post, and it is a fascinating story.
The author starts out with using an existing vulnerability as a benchmark for ChatGPT testing. They describe how they took the code specific to the vulnerability and packaged it for ChatGPT, how they formatted the query and what their results were. In 100 runs only 8 correctly identify the targeted vulnerability, the rest are false positives or claim that there are no vulnerabilities in the given code.
Then they take their test a step further and increase the amount of code shared with ChatGPT so that it also includes stuff of the module that had nothing to do with the original vulnerability. As expected, this larger input decreases performance and also reduces the vulnerability detection rate for the targeted vulnerability. However, in those 100 runs, another vulnerability was described that wasn't a false positive. An actual new vulnerability that the author didn't know about was discovered. Again, the signal to noise ratio is very low, and one has to sift through a lot of wrong reports to get a realistic one, but this proved that it could be used as a useful tool for helping to detect vulnerabilities.
I highly recommend reading the blog post.
As much as I like to be critical about AI, it doesn't help if we put our heads in the sand and act as if it never does something cool.