Speaker frames the discussion in 2025 and cites a Moonshots podcast that reported results from the Ark 1 and Ark 2 reasoning benchmarks. The speaker summarizes headline claims: the benchmarks reportedly identified 44 specific industries or jobs where AI could perform as well as or better than 70% of professionals, and that AI completed tasks "11 times faster" and at "1110 the cost" (phrase spoken). The speaker characterizes these results as both disturbing and freeing and stops short of claiming knowledge work is dead.
Questions this video answers
Ark 1 and Ark 2
44
70% of those in the field
11 times faster
"1110 the cost" (verbatim spoken phrase)
Chapters
Let's, let's take a step back and let's talk about 2025. And you know, for a lot of people, especially if they're not really into artificial intelligence and technology, you know, they're saying, you know, a lot of change, some things change. I still look at YouTube, I still do my social media and things that nature. But when we pull back the curtains a little bit, we talk about some things that have changed dramatically when it comes to artificial intelligence and how it's impacting almost every segment that, that I'm aware of. And one of the most important things, and I'm not going to, I'm not going to belabor a lot of points, but one of the most important things I remember I was listening to a podcast that I love listening to and gather information called Moonshots. And on this they talked about the Ark 1 and Ark to Reasoning test. And that's arc one, arc two. And it's, it's a benchmark to find out how smart AI has become and how really kind of how valuable it is in our day to day. And one of the things that they said, hey, that the difference between an AI and a human is its ability to reason. Well, one of the things in this arc 2 is it does measure reasoning. And so what it did is in the headline of this was Knowledge Work as we know it is dead. And I'm not, I'm not subscribing to that. I'm just going to point out some facts. It said, it pointed out 44 specific industries or jobs where AI could do the job as well or better than 70% of those in the field. So basically what they did is they tested these people and said, hey, if we get a job to AI and we gave it to the, the same person who's a professional. 70% of the time it outperformed that human being and it did it 11 times faster at 1110 the cost. So if you think about that, it, you know, it's almost, it's, it's rather disturbing and, and at the same time it's freeing.
