Your LLM of choice seems to lack world knowledge beyond early 2025, rendering the results pretty uninteresting. What model did you use?
> The comment contains 7 distinct predictions about the future of tech leadership, AI, and hardware. Since the comment is written in 2025 and projects into the future, none of the predictions can be verified as correct yet, resulting in 0 correct. The score reflects a high quantity of specific, non-trivial predictions despite the lack of current hindsight.
I saw a number of score explanations like this. I don’t think it knows what happened in 2025 at all.
dotneter•1mo ago
Thanks. Indeed, scoring 2025 doesn't make much sense, so I removed it. I use Gemini 3 Flash.
beardyw•1mo ago
Disappointing that the comments are not dated, and there is no click through to the original.
But otherwise it's a great idea.
dotneter•1mo ago
There is a link on top [HN thread] to the original post.
subdavis•1mo ago
> The comment contains 7 distinct predictions about the future of tech leadership, AI, and hardware. Since the comment is written in 2025 and projects into the future, none of the predictions can be verified as correct yet, resulting in 0 correct. The score reflects a high quantity of specific, non-trivial predictions despite the lack of current hindsight.
I saw a number of score explanations like this. I don’t think it knows what happened in 2025 at all.
dotneter•1mo ago