This is a review of the reviews, a meta review if you will, but first a tangent. and then a history lesson. This felt boring and obvious and somewhat annoying to write, which apparently writers say is a good sign to write about the things you think are obvious. I felt like pointing towards a thing I was noticing, like 36 hours ago, which in internet speed means this is somewhat cached. Alas. 

Previously, I rode a motorcycle. I rode it for about a year while working on semiconductors until I got a concussion, which slowed me down but did not update me to stop, until it eventually got stolen. The risk in dying from riding a motorcycle for a year is about 1 in 800 depending on the source. 

Previously, I sailed across an ocean. I wanted to calibrate towards how dangerous it was. The forums said the probability of dying from a transatlantic crossing is like 1 in 10,000.

Currently, the people I know working on AI are far more well calibrated to the risk of AI than the general public, and even me, and almost all of them I know think there is more than a 5% chance of an AI catastrophe. That is a 1 in 20 chance, which feels recklessly high.

The thing I wanted to point to was the mental gymnastics I observed in peoples book reviews (if im feeling more contentious I might come back and link to some examples) and the way it made me both disappointed and almost not want to say anything.

I think it's virtuous to note disagreements and it's cool to note agreements, but it's also even cooler and virtuous to avoid misleading people by not saying the trees are not there, when getting into the weeds, literally between the trees. 

There are a bunch of people who are allegedly trying to change the world here. We all allegedly think lots of stuff is at stake. Building a coalition doesn't look like suppressing disagreements, but it does look like building around the areas of agreement. 

If you think there's a 1 in 20 chance it could be so over, it feels to me the part where people are not doing the ‘yes the situation is insane’ even if that is immediately followed up with ‘im more hopeful than them tbc’ is weird. 

On the first day of improv classes they teach you to say ‘yes, and’  instead of using ‘no’, which can kill the scene, when you don't know how to respond, or to move things along. So, yes, and - 

Now is time for the history lesson. The Shanghai Communiqué..

The Shanghai Communiqué was a joint statement issued by the U.S. and China in 1972, breaking a plus 20 year freeze with no diplomatic relations. The communiqué ended a long period of isolation between the two countries and paved the way towards a formal normalization seven years later. 

I think some people critique it for pushing the Taiwan issue down the line, but I like to think about things in the context in which they existed. There was a threat great enough that countries had the incentive to coordinate together. 

These negotiations were happening in the shadow of the cold war, and relations with China were also about counter balancing the USSR. The Sino-Soviet split created space for triangular diplomacy, and improved relations with China, could pressure the Soviets to cooperate on a different arms race. 

I bring up the communiqué because I think it is cool. It acknowledged disagreements, even clearly laid them out, and used some well worded phrases to get it across the line. But the success of the negotiation is, in my read, attributed to the focus on the areas of agreement. Focusing on disagreement would have sunk it before it even started.

It worries me that people who allegedly care about the future going well, and are also at least 5% concerned that AI is not going to go well, are also squandering opportunities to help wake the world up to the dangers that they themselves are seeing, even if they see them slightly differently from the authors.

That said, I am slightly more hopeful than I was yesterday, and hope to feel further more hopeful in tomorrow.