Ever since GPT-3 was unveiled, I've been thinking pretty heavily about increasing my investment in AI-related companies. My first thoughts were to invest in Microsoft and Alphabet (Google) - Microsoft because they are partnered with OpenAI, and Alphabet since they have big AI reseach projects of their own. But in the process of thinking about investing in these companies, I started wondering about the long-term impacts such investments would have on the world - investing in the right or wrong company could dramatically change how the world looks 20 years from now, and whether it is a place I'd want to live in - the worst case scenario would be all humans dead, or even worse; best case scenario is... too amazing to put into words. And then there's plenty of room in between those for how things can go, depending on who makes the important decisions, and how good the decisions they make will be. (While I'm only a single person with modest funds to invest in companies, I also consider that my actions are acausally correlated with those of others sufficiently similar to me, which means the acausal results of any investment I make will be multiplied by an amount that makes my decisions have non-trivial impact on the world).

So the important question is, do I expect Microsoft and Alphabet to do better or worse, in regards to alignment and ethical issues, compared to other actors who will develop AGI in their lieu? (I do expect someone will develop AGI in their lieu) I can think of actors who I expect will likely do worse than Microsoft or Alphabet - the government of basically any country, or firms based in a country with more totalitarian ethics than the US - wheras I can only think of alternative actors who I expect to do roughly as good as Microsoft or Alphabet, but not neccesarily to do better than them. I trust MIRI, but I also don't perceive MIRI as being actively involved in the development of working AI systems; it seems to me that they are laying the important theoretical groundwork for getting things right, but aren't in position to be the ones who actually do the work that needs to be gotten right.

So my main problem here is a lack of knowledge - there almost certainly are other firms who, if I had the relevant information, I would expect would do better on alignment and ethical issues than Microsoft or Alphabet, but I also don't know who those firms are, or why I should expect them to do so. So my question is, for an investor looking to make a AGI-sized profit off of AGI, but also cares about what the future looks like as a result of such investment, what companies will be most likely to result in a good long-term future for humanity?

Note that I'm not asking which company will make the most profit - as long as I reasonably expect that a company will make an AGI-sized profit, that's all I care about on that front. What matters is the impact it has on the desirablity of the future world it will lead to. I'm also not asking about organizations to donate to, because while that is important, that's not the problem I'm chewing over right now.

New Answer
New Comment

1 Answers sorted by

ChristianKl

Sep 22, 2020

70

It seems unclear to me that stock investments in big tech companies have a significant effect on those companies.

Both Alphabet and Microsoft have more then  $100 billion cash at hand at the moment. They are not cash-constraint. 

Thanks for the answer, Christian

If this analysis is correct, that would update me in favour of wanting to invest in smaller, currently under-funded startups that are alignment-aware, over for established players like Alphabet or Microsoft.

Although I'm not completely convinced that this is completely the case- based on the difference in performance between GPT-2 and GPT-3, it seems to me that we are likely in a hardware overhang right now, which means that the question of who is in a dominant position to decide the future of the world via AGI, is largely de... (read more)