Things to think about before you buy - Reflections on Naive Effective Altruism

On Sunday at 13:40 the Director of the Government of AI sent me this very important message.

I've been hesitant to do this because Bob and I have had a rocky relationship ever since he poo pooed my offer to be a guest on his podcast.

Bob's people never did call me.

I hoped that as the two social media giants of so called 'effective altruism', that we would be able to collaborate to do the most good, though I fear that Bob sees me as competition.

But this post isn't about right or wrong.

This post is to illustrate a more important lesson. Too often effective altruists unquestioningly do what they are told by people in fancy sounding positions.

Upon reading Bob's 'Things I Recommend You Buy', I immediately bought all of the products. However, I have now come to see that these recommendations contain basic reasoning errors.

Incorrect understanding of toothpaste

The effective altruism community has long been plagued by poor dental hygiene.

Bob asks:

Did you know there’s a toothpaste that is the same as any other toothpaste, but does a better job at preventing tooth decay and gum disease?

I did in fact know this yes. However, Bob miscorrectly identifies this as being Stannous Flouride toothpaste, when in fact the correct toothpaste is Colgate natural fruit.

Being ineffective and language games (^1)

Bob convenienty recommends ZOMAKE Metal Shoe Horn, as a solution to the shoelaces problem, which in aggregate wastes thousands of hours a year of effective altruists time which could otherwise be used for AI safety.

By his own admission, the ZOMAKE Metal Shoe Horn 'saves [Bob] a few hours a year, making a shoe horn a no-brai

ner.'

Notice the subtle use of language by which he implies that if you don't use his shoe horn you have no brians. If you're intelligent enough to not fall for such cheap language tricks, consider reaching out to me directly to discuss a more effective solution.

Not paying attention to the data

Bob recommends LIFX colour-changing lights 'You can put a big one on your desk, set it to a blue hue, and find it’s just as bright and effective as a winter ‘SAD lamp’.

However, recent research has shown that colour changing lights are unecessary to make you more effective. https://www.medrxiv.org/content/10.1101/2021.10.29.21265530v1

I'm confused by Bob's fixation on colour changing lights. Maybe he was just unaware of the research when he wrote the post? Maybe on his view looking at the data is unecessary, and if you are smart enough you can just use your inside view to work out how to cure SAD?

I contacted the lead researcher directly for recommendations into how to use light more effectively, and I'm very happy with the result.

Confusing Symptoms and Causes

Rob recommends many ways of getting better sleep - melatonin, a fan for sleeping, HOMMINI Sleep Mask, Pitrok Crystal Deoderant.

This confuses cause and effect, symptoms and causes, the cart and the horse. Sleep is a read herring. The key question to ask is why do we want better sleep? To have more energy of course!

I attack this problem directly, using nootropics.

Bill's only good suggestion here is to take melatonin, but he gets the reasoning entirely backwards.

Using a conditional imperative, he says: 'If you find it makes your dreams too vivid, or otherwise interferes with your sleep, start by taking a lower dose and build up gradually.'

But as I've written about previously, dream vividness is a primary determiner of your quality of life - you should take melatonin in order to have vivid dreams!

The fan is also useful as a bedside table.

Naive understanding of option value

Bob recommends Kirkland Signature's Minoxidil Foam to prevent hair loss, and then immediately afterwards recommends Wahl Lithium Ion Stainless Steel Groomer, seemingly unaware that these two recommendations are in direct contradiction with each other.

(I confess, I too was seduced by his compelling logic, and only realised the inherent contradiction too late)

Naive understanding of information security

Rob recommends U2F security keys, a classic example of trying to impose technical solutions upon non-technical problems. Here's how I approach information security:

1. If I need to talk about someone else, I will use a pseudonym to protect their identities, such as 'Bob' for 'Robert'.

2. I use the same password for everything as recommended by https://1password.com/

3. I understand the trade offs between being safety and risk. For example, I don't lock the front door to my house and haven't done for 5 years.

4. I never reveal personal information about myself or others online.

Credit where credit is due

as Robert Pirsig puts it, “The world’s greatest fool may say the Sun is shining, but that doesn’t make it dark out.”2

We must give credit where credit is due, and to Bob's credit his recommendation to use chopsticks was very good.

This begs the question, when is credit due?

In his recommendation of neoliberal credit cards he only mentions once that he will get $100 if you open an account using the links he suspiciously provides.

To maintain my integrity as an independent researcher I don't use amazon affiliate links or tell you to take sign up for credit cards in order to give me money.

If you like my work please give me money directly on Patreon.

Footnotes

^1