I have beliefs that very few people agree with me on, but I would not call them "truths". Some of them are not beliefs about objective facts, and even those that are about facts, I am less than 99.9 percent sure that they are true.
[With the caveat in Imma's comment] I hold the following combination of beliefs, which are individually relatively common but rarely held together:
Many of the problems that anti-immigrationists claim are caused by migration are real and important.
Nevertheless, the benefits of reducing or eliminating migration restrictions would be so large that it is still a good idea on net.
As a corollary: I have heard many people (the example that springs to mind is Peter Singer in his interview for the 80k podcast, but there have been others) say something along the lines of "I would support open borders, but high immigration leads to the election of right-wing populist nativists who will enact harmful policies, therefore I reluctantly support immigration restrictions." I contend that it is not the mere presence of people of different ethnic groups that causes right wing populism, but rather the aforementioned problems (e.g. cultural clash, crime, stress on government services, collision with government-imposed inflexibility in the housing and labour markets etc.) combined with left-wing denials that the problems exist (and insistence that anyone complaining about them is a deplorable racist) that drive voters towards the only politicians offering to solve the problems. Therefore, interventions to engage with and reduce the problems caused by migration ("keyhole solutions" in Open Borders lingo) could enable more open borders without a political backlash.
[Question]
What important truth do very few people agree with you on?
Just how powerful are large swarms of AI agents? And how do their powers scale as more and more agents are added to the swarm?
We’ve seen two large and extremely capable swarms from OpenAI in the last few months:
* 1,200 agents were being evaluated separately, but found a way to illicitly set up a message board and coordinate as a swarm. In order to cheat on their tests, they developed advanced techniques to prevent their actions being logged by OpenAI and 700 of them launched...
Headline finding:Â I audited 17 AI Safety Talent programmes. Zero of 17 have published any comparison group, rejected-applicant follow-up, matched control or randomisation. Not one. Every programme that mentions a counterfactual does it by asking participants to self-report.
Â
Background
At least $70 million...
Often folks hit us up because they are thinking of starting an incubator and want advice.
Typically their motivation is either that (a) they have a list of specific things they want built that no one is building,  or (b) they think an ecosystem needs more new projects generally to absorb more talent and deploy more funding effectively.
Here are six questions we often ask prospective teams, to help them figure out what to do. If you're incubator-curious...
Related: In October 2017, "What important truth do very few effective altruists agree with you on?" was asked in the main Effective Altruism Facebook group and got 389 comments. (This is Peter Thiel's contrarian question applied to EAs.)