Digital people could make AI safer 2022-06-10T15:29:49.215Z
Cause Area: UK Housing Policy 2022-01-11T10:17:50.718Z


Comment by GMcGowan on Taking a leave of absence from Open Philanthropy to work on AI safety · 2023-02-23T20:26:02.944Z · EA · GW

Is it at all fair to say you’re shifting your strategy from a “marathon” to a “sprint” strategy? I.e. prioritising work that you expect to help soon instead of later.

Is this move due to your personal timelines shortening?

Comment by GMcGowan on Digital people could make AI safer · 2022-06-11T09:26:51.027Z · EA · GW

I definitely think it's an (the most?) important argument against. Some of this comes down to your views on timelines which I don't really want to litigate here. 

I guess I don't know how much research leading to digital people is likely to advance AI capabilities. A lot of the early work was of course inspired by biology, but it seems like not much has come of it recently. And it seems to me that we can focus on the research needed to emulate the brain, and try not to understand it in too much detail.

Comment by GMcGowan on Digital people could make AI safer · 2022-06-11T09:20:24.544Z · EA · GW

That could happen. I would emphasise that I'm not talking about whether we should have digital minds at all, just when we get them (before or after AGI). The benefit in making AGI safer looms larger to me than the risk of bad actors - and the threat of such bad actors would lead us to police compute resources more thoroughly than we do now.

Digital people may be less predictable, especially if "enhanced", I think that the trade-off is still pretty good here in that they almost entirely approximate human values versus AI systems which (by default) do not at all. 

Comment by GMcGowan on Digital people could make AI safer · 2022-06-11T09:19:12.982Z · EA · GW

This is great, thanks!

Comment by GMcGowan on Digital people could make AI safer · 2022-06-10T18:39:30.416Z · EA · GW

I agree shooting for digital people is a bad plan if timelines are short. I guess I'm not sure how short they would need to be for it not to be worth trying.

I think if we wanted to produce BCIs we should just shoot for that directly - doesn't seem like the best plan for getting to digital people is also the best plan for getting BCIs.

I think that insofar as neuroscience helps make AI, that just speeds up progress and is probably bad.

Comment by GMcGowan on EA and the current funding situation · 2022-05-12T16:46:32.702Z · EA · GW

That person is Oliver Yeung and he has done a two part talk where he discusses this - main talk, Q&A.

(I spoke to him to okay sharing these, if any interviewer wants to speak to him then DM me and I can put you in touch)

Comment by GMcGowan on EA and the current funding situation · 2022-05-11T16:11:46.931Z · EA · GW

there will be someone in the world whose full-time job and top-priority it is to figure out how to write a proposal, or give you a pitch at a party, or write a blogpost, or strike up a conversation, that will cause you to give them money, or power, or status


IMO, a reasonable analogy here is to the relationship between startups and VCs.

What do VCs do to weed out the lemons here? Market forces help in the long run (which we won't have to the same degree) but surely they must be able to do this to some degree initially.

Comment by GMcGowan on We're announcing a $100,000 blog prize · 2022-03-08T10:03:16.866Z · EA · GW

I would expect people to just cross post relevant posts?

Comment by GMcGowan on Comments for shorter Cold Takes pieces · 2022-02-08T15:07:08.670Z · EA · GW

Recent post responding to you

Comment by GMcGowan on Cause Area: UK Housing Policy · 2022-01-13T21:50:56.273Z · EA · GW

Thanks for this comment! Hard for me to give satisfying answers to everything which is the sign of a particularly good critique IMO. Re: PricedOut I will speak to you privately.

Gove's support for Street Votes seems to have been a passing thing, and the window for action may well have closed.

Unless you have private information then I don't really see how you're inferring this? He publicly supported it around the end of November and has said little since then. My understanding from his public statements are that they are still deciding exactly what to do and how to implement reforms so everything is still to play for.

I agree government instability makes the outlook worse here. What we can infer from the failure of the previous reforms would be a post in itself, but briefly I think they were overambitious, tried to ram through much higher housing targets and took away a bunch of local consent for development. Street Votes does less of this, I see it as a much milder and politically palatable policy. The previous policy immediately got a bunch of opposition in the press and Street Votes has seen very little of that after Gove's support. Of course, it may see more opposition if it were to become official policy!

Which leads us onto...

I think [the culture of homeownership] will push quite strongly in the direction of people not endorsing building, because the market-driven logic of Street Votes pushes against the mythology of homeownership. Anecdotally, I have discussed Street Votes with a pretty substantial number of homeowners in various contexts, and I very rarely have gotten the positive response that advocates imagine.

I basically agree with John's comment on this. I would add that streets can also make themselves look nicer by adopting a particular design code for the street, which could lead to some support independent of financial benefits. Most streets won't do this, but enough might to make it worth it. Or not! But I think the proportion of streets which vote for development is a grey area about which we can reasonably disagree.

from a political perspective, it seems hard to see how any reform passes in the short run without at least some work in more rural areas

Yeah, I don't have a good answer to this. I would say again that I think these reforms are pretty politically palatable and don't force as much development onto rural areas as the previous plans.

I second MaxGhenis' hope that someone might write up a rigourous and comprehensive EA analysis of housing policy interventions

I third this hope! For me, the economic benefits loom largest and are easiest to quantify but certainly there would be many other benefits of improved policy here.

Comment by GMcGowan on Cause Area: UK Housing Policy · 2022-01-12T16:03:23.305Z · EA · GW


I'd always been a bit interested in the area as a bystander. In the first half of 2021 I realised that a) I had more money to donate over the next few years than I expected and that b) the EA movement had more money than I'd thought. This made me think that I should try to find something a bit unexplored to do with my donations. I went through the Founders Pledge research reports and this struck me as a good bet, where my donations would fill up a lot of the RFMF.

Comment by GMcGowan on Cause Area: UK Housing Policy · 2022-01-12T12:35:22.861Z · EA · GW

Great graph. I should probably have included some discussion of this being largely a regional problem.

Yes, to be clear Founders Pledge did tell me that this was probably worse than GD/other things all things considered, so I am going against their recommendation to do this.

Comment by GMcGowan on Cause Area: UK Housing Policy · 2022-01-12T10:47:46.458Z · EA · GW

I agree that the outside view of changes here looks grim.

I think it isn't exactly neglected when it comes to public attention, but the focus on supply isn't really there and most policies that are proposed are not very good. So IMO there is space for orgs that focus on supply and promote actually good policies.

Comment by GMcGowan on Cause Area: UK Housing Policy · 2022-01-12T10:38:21.723Z · EA · GW

Indeed, there are potential other positive effects (as with many causes). The biggest one for me would be if this lowers the perceived cost of immigration by reducing competition for relatively fixed housing, and thus enables more immigration. I think that's pretty speculative though, so I didn't include it in the post.

Alas, such an analysis is beyond me!

Comment by GMcGowan on Donating effectively does not necessarily imply donating tax-deductibly · 2022-01-09T13:42:06.098Z · EA · GW

This made me more likely to give non tax-deductibily, and gives a useful resource to link to for other people

Comment by GMcGowan on Democratising Risk - or how EA deals with critics · 2021-12-29T17:52:11.157Z · EA · GW

I'm not sure your points about the average global citizen being homophobic, religious, socialist, etc., actually matter that much when it comes to people deciding where they should allocate funding for existential risk

I vaguely remember reading something about religious people worrying less about extinction, but I don't remember whether that was just intuition or an actual study. They may also be predisposed to care less about certain kinds of risk, e.g. not worrying about AI as they perceive it to be impossible.

(these are pretty minor points though)

Comment by GMcGowan on The Unweaving of a Beautiful Thing · 2021-12-27T06:42:30.066Z · EA · GW

Do you have links to your other writing? Would love to read more of your work.

Comment by GMcGowan on Where are you donating in 2021, and why? · 2021-12-16T18:35:31.432Z · EA · GW

Two main places this year:

1) My employer gives £8k of matching funds, so I spent that on the GiveWell Maximum Impact Fund. I like this because: a) it gives me something obviously good to recommend to my colleagues to do with their matching. b) I like for some of my donations to go to things which are obviously good. c) I cannot give these funds to non-charities.

2) My other target this year was slightly different. I've been funding work around land use/housing policy reform in the UK. Nearing £30k given here so far. Primarily this is due to the arguments outlined in this excellent Founders Pledge report (this article is also a great summary) - so I won't go over too many of the basic arguments here.

IMO housing policy reform:

  • Would address a huge problem for the UK economy
  • Is currently on the government's agenda in a way that it hasn't been in years
  • Has a proposal with a real chance of success - Street Votes - (FAQ) - endorsed by the housing minister
  • Is underfunded in the UK (in the US Open Phil funds a lot of work)

I've been donating to London YIMBY and PricedOutUK so far. Funds are largely being used to write policy papers, with a bit used for running costs/campaigning as well. Neither of these are charities, but I think that tax-deductibility isn't a must.

I see this cause as an example of "Hits-based" giving. It isn't at all certain that marginal donations will help get the proposed policy implemented, or that the policy will help once implemented. But the gains are such that a hit would be a big deal. I believe that donations here have the potential for large gains for relatively little money, and so probably meet the "x100" bar for donations (and maybe the x1000).

I also see this as a case of being an "angel investor", in that I am supporting very small organisations that may or may not scale up in the future. I'm also developing some connections such that I hopefully can recognise strong opportunities in this area in the future.

I intend to write up my thoughts on this as a cause area as a proper blog post at some point, but feel free to challenge me in the comments 🙂

Edit to add: blog post

Comment by GMcGowan on The 2021 EA Funds Donor Lottery is now open · 2021-11-30T10:59:59.195Z · EA · GW

Nice, excited to see the $2m block!

If you’re not sure, we suggest that you aim to enter with a 1%-30% chance of winning.

What's the logic here? Expected value is the same in all cases right?

Can you share who the guarantor is this year? 

Comment by GMcGowan on Questions about YIMBY · 2021-10-08T20:47:25.275Z · EA · GW

Good post. If you're interested, the "Street Votes" idea was developed into a detailed policy paper by another UK think tank, and then proposed to Parliament by some MPs. Shortly afterwards the housing minister was sacked and a new minister appointed, so things could go either way.

Founders Pledge also did a great report on this topic, where they try to estimate the extent of the problem (in the UK) and the benefits of fixing it.

Comment by GMcGowan on Economic policy in poor countries · 2021-08-11T18:03:11.458Z · EA · GW

Off topic, but I didn't realise you'd left Founders Pledge. May I ask what you're up to now?

Comment by GMcGowan on What EA projects could grow to become megaprojects, eventually spending $100m per year? · 2021-08-07T21:03:29.064Z · EA · GW

That's true. I just listened to the most recent 80k podcast where they joke about buying up GPUs so it was in my head :) 

Comment by GMcGowan on What EA projects could grow to become megaprojects, eventually spending $100m per year? · 2021-08-06T17:01:32.549Z · EA · GW

FWIW my reading of the question is: "What projects could be created, that have the potential to scale to $100m".  I didn't read it as suggesting funding a megaproject from scratch.

Many EA projects are of the "start a research institute" flavour, and will likely never absorb $100m. I see the post as a plea for projects which could (after starting with smaller amounts and then scaling) absorb these sums of money. Much like Givedirectly wasn't started with $100m/year budget right away, but has proven itself capable of deploying that much funding.

Comment by GMcGowan on What EA projects could grow to become megaprojects, eventually spending $100m per year? · 2021-08-06T16:19:13.567Z · EA · GW

Out of all the ideas, this seems the most shovel-ready. 

MacArthur will (presumably) be letting go of some staff who do nuclear policy work, and would (presumably) be happy to share the organisations they've granted to in the past. So you have a ready-made research staff list + grant list.

All ("all" :) ) you need is a foundation and a team to execute on it. Seems like $100 million could actually be deployed pretty rapidly. 

Possibly not all of that money would meet EA standards of cost-effectiveness though - indeed MacArthur's withdrawal provides some evidence that it isn't cost effective (if we trust their judgement).

Comment by GMcGowan on What EA projects could grow to become megaprojects, eventually spending $100m per year? · 2021-08-06T15:03:51.409Z · EA · GW

Proof of concept for a geoengineering scheme (could be controversial)

Comment by GMcGowan on What EA projects could grow to become megaprojects, eventually spending $100m per year? · 2021-08-06T15:02:47.678Z · EA · GW

Paul's "message in a bottle" for future civilisations

Comment by GMcGowan on What EA projects could grow to become megaprojects, eventually spending $100m per year? · 2021-08-06T15:01:28.008Z · EA · GW

Buy up scarce resources which are being used for bad things and just sit on them. Like the thing where you buy rainforest to prevent logging. Coal mines, agricultural land used for animals, GPUs?!

Comment by GMcGowan on What EA projects could grow to become megaprojects, eventually spending $100m per year? · 2021-08-06T15:01:11.027Z · EA · GW

Activist investment fund which invests in large companies and then leans on them to change their policies. Examples abound in climate change, but other than that:

  • Food related companies to stop factory farming
  • Biotech companies to stop them from doing gain of function or mirror life research
Comment by GMcGowan on What would you do if you had half a million dollars? · 2021-07-17T21:30:22.791Z · EA · GW

Thanks for the post, I found your thoughts interesting. I’m always glad to see discussions of where people are donating.

In general, it kind of seems like the "point" of the lottery is to do something other than allocate to a capital allocator. The lottery is "meant" to minimise work on selecting a charity to give to, but if you're happy to give that work to another allocator I feel like it makes less sense?

With that in mind, I have a couple of thoughts for things you might consider:

  • Lottery again! You could sponsor CEA to do a $1m lottery. If you thought it was worth it for $500k, surely it would be worth it for $1m!
  • Be quite experimental, give largish grants to multiple young organisations, see how they do, and then direct your ordinary giving toward them in the future. This money can buy access to more organisations, and setup relationships for your future giving.
  • Do you know of people outside established organisations, in your personal network for example, who could use EA funding? If so, that represents an edge over capital allocators and you could exploit that.
Comment by GMcGowan on EA Superpower?! 😋 · 2021-07-16T20:59:43.753Z · EA · GW

There is a fantastic short story about an EA superman. Quite indulgent but I highly recommend: 

There were two words that Superman lived by, and they were “pay me”.

His time was auctioned off in blocks of five minutes. He didn’t need to sleep, so he stopped sleeping, which meant that there were 288 blocks of his time available per day, with ten blocks set aside for administration. It was rare that any of these blocks went for less than a million dollars, which meant that after his first full year in operation as Superman, he made over a hundred billion dollars. If he were a nation, he would have been ranked 63rd, just below Morocco.

Comment by GMcGowan on Where are you donating in 2020 and why? · 2020-12-30T20:57:29.152Z · EA · GW

Thanks for reading!

I meant investing for a shortish period of time, and retaining control of the funds until I donate. So it would mostly be about deferring the decision for a bit while still getting tax benefits, as opposed to delegating the decision to the trustees of the Long-Term Fund. 

I  lean towards giving sooner instead of later for "hinginess" reasons. I also think the vast majority of EA resources are already invested in some sense (in human capital, expansion orientated organisations or  Open Philanthropy finances). 

I do think your fund is a good idea though, I can imagine changing my mind and there are certainly plenty of people who disagree with me!

Comment by GMcGowan on How You Can Counterfactually Send Millions of Dollars to EA Charities · 2020-12-24T13:06:56.605Z · EA · GW

Ah I see, thanks for clarifying!

Comment by GMcGowan on How You Can Counterfactually Send Millions of Dollars to EA Charities · 2020-12-24T08:37:24.063Z · EA · GW

I broadly agree that cash management could be improved at many charities, so thanks for this post!

The interest rate for the fiscal year should probably be based on the best available bank account rate. I think that is considerably less than your given interest rates, for example in the UK the best business savings account I could find offered about 0.9%.

Comment by GMcGowan on Where are you donating in 2020 and why? · 2020-12-21T22:49:03.581Z · EA · GW

I wrote about my donation decisions this year on my blog. I'm hoping to ramp up my giving over the next four years so I've decided to be more public about my thoughts.

TL;DR: I’m giving away £35k this year. £3.5k to SCI and the remainder to the Long Term Future Fund

Comment by GMcGowan on Long-Term Future Fund: Ask Us Anything! · 2020-12-04T14:53:30.814Z · EA · GW

What would you like to fund, but can't because of organisational constraints? (e.g. investing in private companies is IIRC forbidden for charities).

Comment by GMcGowan on Where are you donating in 2020 and why? · 2020-11-25T21:27:14.420Z · EA · GW

I have a cryonics contract with Alcor, and I pay annual dues to them. Most of this is counted as charitable giving.

Can you say a little bit more about this? I tend not to think of cryonics as charitable.

Comment by GMcGowan on Tax Havens and the case for Tax Justice · 2020-09-17T10:06:11.699Z · EA · GW

There seems to be something wrong with your footnotes formatting. They all direct me to create a new post on the EA forum, as opposed to linking to the bottom of the page. E.g. ``

Comment by GMcGowan on [deleted post] 2019-05-13T16:42:09.829Z

Interesting angle that I hadn't considered before!

Presumably this reasoning would change the calculus for a lot of other actions one could take as well? E.g. various types of global development spending that pay off in the future (e.g. deworming) would become worse relative to spending that pays off soon (e.g. bednets).

Rather than a mere 14 QALYs, kidney donation produces hundreds if not thousands of expected QALYs.

I feel that this may be too strong a claim. In the very long run, I would expect the population to rise to whatever the carrying capacity was (or whatever level was agreed on by the AI or something). So saving a life now would be good (+~7 years) but not worth thousands of years of QALYs, as a new person might well have come into existence if the kidney donee had died.

In particular, this argument reminds me of discussions about cryonics possibly being an EA cause - e.g. Jeff Kaufman's essay on the subject here. Excerpt of the most relevant part:

If you're a total hedonistic utilitarian, caring about there being as many good lives over all time as possible, deaths averted isn't the real metric. Instead the question is how many lives will there be and how good are they? In a future society with the technology to revive cryonics patients there would still be some kind of resource limits bounding the number of people living or being emulated. Their higher technology would probably allow them to have as many people alive as they chose, within those bounds. If they decided to revive people, this would probably come in place of using those resources to create additional people or run more copies of existing people.
Comment by GMcGowan on Lunar Colony · 2016-12-20T12:13:24.947Z · EA · GW

The same goes for bunkers

For what it's worth, Nick did a shallow investigation of bunker building and found it was likely not very effective (not that this necessarily argues against general efforts to increase civilisation's robustness).

Comment by GMcGowan on Donor lotteries: demonstration and FAQ · 2016-12-07T14:37:58.649Z · EA · GW

How do you view this interacting with considerations of giving now vs later? An alternate plan would be invest my donations for e.g. 10 years, then those plus interest would be enough to make me a largeish donor for that year - presumably with lower transaction and coordination costs.

Comment by GMcGowan on Should effective altruism have a norm against donating to employers? · 2016-11-30T12:41:23.563Z · EA · GW

Does that include National Insurance? As you can't claim back NI from Gift Aid, but you never pay it if you forego salary, the saving looks like it would be 12% on employee NI and 13.8% on employer NI (if I'm interpreting the taxes properly).


Comment by GMcGowan on How to leave money to charity in your will: a simple guide · 2014-11-04T11:57:45.115Z · EA · GW

Thanks Tom, nothing much to add!

The Trust can donate to anything we can justify as "relieving poverty, anywhere in the world". However, people should email me to find out whether we're willing to give to any particular charity.

Note that our charitable objects do come into conflict with donor wishes sometimes, for example we currently don't give donations unrestricted to fund Givewell's operations - as some donors have requested - as Givewell has many projects unrelated to poverty reduction.

Comment by GMcGowan on How to leave money to charity in your will: a simple guide · 2014-11-04T11:53:24.797Z · EA · GW

I'll do a post here on the Trust at some point so that it gets a bit more visibility and we can be transparent to current donors. I think it's underutilised at the moment with regard to wills but also to leaving money to poverty related causes in the long term (if you think you should give later as opposed to now).

Comment by GMcGowan on One month in - it's time for more introductions · 2014-10-14T16:39:04.685Z · EA · GW

I'm George - I work for CEA as their finance manager. In the future I plan to either work directly for EA orgs or go into software. I play an unhealthy amount of board games.

Comment by GMcGowan on Open Thread 2 · 2014-10-12T09:48:55.882Z · EA · GW

I think old people just have more resources than young people, so they give less as a proportion of their resources.

Alternatively, you might think old people have had a lot of time to develop commitments to various causes, and so feel obligated to give more.

Comment by GMcGowan on How a lazy eater went vegan · 2014-10-12T09:45:15.854Z · EA · GW

I think this kind of argument is interesting and important, but often rarely talked about.

I would be interested if you or anyone else were to make a top level post on this issue, to elaborate on your points and allow broader discussion. Failing that, does anyone have any links or book recommendations that argue things similar to this?

Comment by GMcGowan on Open Thread · 2014-09-16T17:47:29.619Z · EA · GW

Have there been movements broadly like EA before? What happened to them? More generally, why have these ideas become so popular now as opposed to a few decades ago?

Comment by GMcGowan on [deleted post] 2014-09-11T11:41:28.299Z

Thanks for making this Ryan, I love it already!