Sam Glover (fka Atis)
@sam_atis
Followers
2K
Following
9K
Media
111
Statuses
3K
If you are interested in social science or effective altruism, read https://t.co/LCWUcBUQCE. I do forecasting research.
Joined June 2021
I bought a guy in Japan a beer two years ago and he insisted on taking my details to pay me back. Seems like he struggled to set up a PayPal but kept working on it and finally managed it! Paid back with interest.
98
342
15K
Claude’s advice for how to get a better deal when looking for a flat is… unusual.
43
102
5K
Found a guy on TikTok who livestreams replications and it’s a real social science blackpill. He’s doing basic checks in google sheets and finding obvious errors or that findings have been portrayed in ways that are extremely misleading.
21
197
5K
A friend of mine was complaining that everyone advises her to invest in Vanguard, but she bought a load of stock earlier this year and has lost almost half her savings. Turns out she was invested in an agrochemical company called American Vanguard, which has halved in value YTD.
35
71
2K
At what point does it become genuinely kinda mean to keep pointing out how poor we are lol? ATM British people are fair game to make fun of, but presumably at some point it's just people in a rich country laughing at people in a much poorer country for having no money.
187
53
1K
The head of pandemic preparedness in the UK will literally make less than most random 24 year olds I met when I was in NYC lol
17
60
1K
It's so funny to me that a friend of a friend wanted to get rich, read 'business for dummies', and then created a company that buys and sells medical equipment and makes him millions per year.
9
34
1K
Main lesson from travelling was that basically every young person thinks their own country is uniquely fucked. From New Zealanders to New Yorkers and from Spaniards to South Koreans, ‘you don’t get how bad things are for us’ is everywhere.
18
20
504
It's crazy that we have an an alien form of intelligence that is >90th percentile ability in many skills, that knows more about the world than any human, that is willing to talk with you and work with you at any hour of the day, and some people still think it's a marketing con.
Newsletter: Gen AI is a marketing con perpetuated by Sam Altman and Dario Amodei through cynical marketing half-truths carried by a tech media that fails to ask the right questions. Outside ChatGPT, generative AI companies don't get much traffic at all.
25
28
528
Moments after an Effective Altruist discovers he has an extremely attractive third cousin:
11
16
458
@stuartbuck1 I think his TikTok account is lthlnkso (with Ls instead of Is), the name is ‘Quick Thoughts’.
5
2
362
In a way, I'm happy that Americans still feel it's fair game to make fun of UK GDP growth and europoors. When people start saying it's punching down that we'll really have a problem!.
13
7
330
This is a big problem in EA that someone is finally shining a light on, I have had to tell like ten people not to do this.
13
14
286
Proper anti-cheems mindset stuff. 'Do you really think you're going to start a successful business after reading a book called Business for Dummies?'
2
4
252
Long-term nuclear waste warning messages are very eerie.
22
18
262
Possibly the worst advice ever on how to give to charity comes from the FT's 'How to Give It' spread, which seems to be focused on how to donate in such a way that you personally get social status and/or luxury goods in return.
7
14
261
(She found it very funny when this was pointed out btw, it’s not some sad situation where she desperately needed the money).
1
2
234
@arpitrage @s8mb In the UK it’s fairly common to switch out ketchup for mayo on your fries if you’re not a fan of spicy foods.
24
9
210
Suppose you travel to the year 2100 and meet your great-grandchild, and they tell you: ‘You should definitely end your involvement with Effective Altruists. EAs end up doing a lot of harm’. What is the most plausible explanation for his/her comment?.
74
5
169
It’s sad to see our American friends continue to make these jokes when their liberty is now far more at risk than ours. Like a German continuing to make jokes about the UK’s economic growth. Would’ve made sense a decade ago, less so today!.
I've ruined the vibe at two separate happy hours by asking a Brit if they think they live in a free country. Thought my puckish demeanor would charm. Will workshop the delivery further.
5
10
161
I think more journalists should consider that they are actually serious about creating superintelligence, or that this is actually possible, rather than just defaulting to "cool story to keep the cash coming in".
10
14
150
I don’t think there is currently a nicely written 1000 - 5000 word piece about AI as an x-risk that I can send to non-EA friends that will make them say ‘Oooh, now I get it’. Is there a go to piece (or podcast or something else) that I just don’t know about?.
27
3
138
It's kind of weird that stuff like this never comes up in the discussions about the alleged crisis of masculinity. Men are fighting less (and drinking less) and domestic violence rates have fallen significantly. Seems much more important than # of Andrew Tate Youtube views.
Violence inflicted on the weekend has fallen by half over the past decade in Britain. What accounts for this drastic shift? .
3
9
144
A quick reminder that Dominic Cummings' view in 2021 was that if Labour kept Starmer as leader and the Conservatives switched leader, the chance of a Labour majority in the next GE was under 5%. We'll see what happens.
16
22
120
Suppose you travel to the year 2100 and meet your great-great-grandchild, and they tell you: ‘You should definitely end your involvement with Effective Altruists. EAs end up doing a lot of harm’. What is the most plausible explanation for his/her comment?.
57
4
127
It should be easier to get drugs like Methylphenidate (Ritalin) or Adderall, which make you much more productive with very little cost.
10
4
127
The British public seem to be just as worried about x-risk from the bees dying out as they are about x-risk from AI. This is not because they are worried about neither. 17% of people think the bees dying out is likely to cause human extinction, and the same % think AI is.
9
15
113
Me: Toby Ord has claimed that the world spends significantly less money on reducing extinction risk than on ice cream. My grandmother: That doesn't seem particularly surprising at all. I'm more surprised that you would think I would be surprised by that.
2
3
105
I don't really understand why EAs believe that the best PR always comes from being honest and direct. You may want it to be the case that honesty is the best policy, but I don't think it's true. Think of all the massively successful companies that are often dishonest.
27
1
92
Calling Tyler Cowen 'alt-right' is so absurd that it indicates either a total ignorance of what 'alt-right' means or a willingness to defame someone for seemingly zero benefit.
1
1
78
There’s a strong case that degrowth is basically not important as an idea, so it’s a waste of time to write articles criticising it. Ideas that are actually inimical to growth (e.g. making building illegal) are not popular because of the success of ‘degrowth’ as an idea.
Curious!. Degrowth cannot work, and yet Nature publishes that it can. What does this tell us about Nature?.
8
2
79
‘Bets are a tax on bullshit’ also ends up meaning that your explicit bets can be quite a different vibe to your abstract claims. I know a few superforecasters who are nutty when making general claims about the world and then meticulous when making objective forecasts.
Are you a Bryan booster? A Caplan critic? .Come predict whether he wins his current slate of bets:
3
6
80
X is a fundamentally different platform to Twitter, and it’s so much worse.
11
4
67
Every day I still feel completely in awe of the LLMs that I use. It's absolutely amazing that people have created these. It's nuts that there are so many people who never use LLMs and don't see what the big deal is, it's so obvious that these are going to change the world!.
3
2
72
I was in a conversation with a fairly well known British political pundit this evening and said ‘I’m not really sure what you’re saying is true, would you be willing to bet on this?’. Reaction was weird, instantly very cold and ‘Sorry, I have to go and talk to someone else.’.
2
2
74
@gbrl_dick PS5 is an example where this wasn’t true at all! There were huge supply issues and it was impossible to get one for a year (years?) unless you paid a massive premium.
9
0
71
@Techerous The first stock I bought when I was like 18 was dominos because it felt cool buying a pizza from a company that I owned a piece of.
0
0
61
I think stated preferences are underrated (and revealed preferences overrated) among people I know. If someone says ‘I use this app all the time but I hate it, I feel addicted, and I wish it didn’t exist’, they’re probably right that it doesn’t provide them much value.
4
4
69
I am very sympathetic to the view that trawling through one academic’s work to find any mistake you can is unreasonable. But it’s a little worrying when the response is ‘You think that’s bad? All social science papers have huge methodological problems!’.
3
6
66
Has anyone ever written a post making the case that a valuable thing to do if you're running a non-profit is try extremely hard to get non-EA funding before you try to get EA funding? Seems like that's probably the case, but I don't think I've heard anyone emphasise it.
10
0
60
Boomers are right on this one - spending like £4 to avoid sitting in the cold all evening is obviously worth it.
I've always been confused how the avg energy bill is like £180 when mine rarely goes above £60 but each Xmas I am reminded that boomers love to turn their homes into saunas and not to sound like Lee Anderson but I think sometimes you should just put a jumper on.
4
0
57
A friend and I had a discussion a few years ago about phrases you could use in an online argument that would annoy people the most, and we decided ending every reply with ‘think before you post’ would be the most effective. Funny to see it being used in government comms!.
Think before you post.
3
0
58
@cremieuxrecueil @trussliz What specifically do you mean by this? If you’re referring to a comeback in UK politics, I would be happy to bet with you at the right odds.
1
0
54
I think I overestimated how senior this position was on the basis of the role title and the 'senior' tag on 80,000 Hours (which has now been changed). I still think the salary is low but it's less bad than I thought.
@EgeErdil2 @sam_atis @JacquesThibs The UK role posted above isn't even close to the most senior, it's middle management. Executives in DHSC/NHSE make more.
2
0
50
Buying a new domain to upload a single, long essay to a beautiful website is the new meta for having a big impact. Might get nerfed soon - get yours up ASAP!.
A NARROW PATH. If attempts to build superintelligent AI succeed, we face extinction as a species. We must choose a different path, one where we control what we create. One where AI is a tool for human advancement, not a successor species. Humanity has no plan, so we built one.
5
1
53
@GeorgistSteve Or just move mostly to exams and have a few essays where people are encouraged to use GPT. I'm not totally sure a keystroke recorder would work, seems like you can just have your essay pre-written on another laptop and then write it 'organically' on the one you use to submit.
7
1
48
1. Buy and hold ETFs. 2. Use Anki to do well in university exams (and to memorise anything you want to). 3. If you succeed in most things you attempt, be more ambitious. 4. Read the book ‘Superforecasting’. 5. Use the Hostelworld app when you travel.
4
1
48
Seen a lot of people endorsing this so I may as well say I think this is basically wrong and that good PR is important and takes more effort than ‘just be honest and act with integrity’.
8
1
48
1
0
46
I don’t think this is very likely.
What I fear could happen in October:. The Democrats will lose the White House unless they come up with a truly game-changing 'hail mary' strategy. Given their willingness to demonize their opponents as 'violent insurrectionists', one logical option would be to stage a false flag.
2
1
43
This thread is worth reading to understand how MacAskill’s appearance on the Sam Harris podcast has been received by Harris’ fans.
7
2
44
“Should you donate to Global Health and Development charities or Animal Welfare charities?”. “Both”. “You can’t choose both”. “I just did.”
1
0
43
One thing I noticed in NYC is that Americans (or those in New York, at least) are obsessed with this sort of thing much more than people in London. Certain cars, watches, suits - everyone seems to know what they’re intended to convey in ways that I’ve never thought about.
I’m a woman. I worked in luxury watch resale for several years. ‘Cheapest I could get away with’ Business School status watches make you look like you aspire one day to ruin a rather boring marriage by having frankly mediocre sex with a dumpy cakey foundation early 20s secretary.
8
0
42
I've heard all your boring suggestions for how to improve GDP growth: build more housing, open the borders, re-join the Single Market. The real solution is to give everyone amphetamines.
0
4
39
Reactions to this seem pretty crazy to me, seems apparent that Caplan is basically right here so long as it’s true that a somewhat meaningful proportion of Americans are not up to date with their tetanus shots.
My doctor talked to me like a child when I refused a tetanus booster. This disease kills about 2 Americans per YEAR!.
7
0
39
On the other hand, I suppose this is at least optimising for something, it's just something I don't think is important. Maybe I have to begrudgingly respect the optimisers of all stripes.
4
0
40
How should the EA Forum handle questions like these? It’s a sincere question from someone who is genuinely interested in EA and wants to do good. But it’s also low effort and it’s pretty obvious that interventions to prevent school shootings won’t be particularly cost-effective.
15
1
38
@Kirsten3531 She had only recently started saving/investing so it isn’t as bad as it sounds in the original tweet tbh.
0
0
39
It doesn’t seem like a sober and true take to me. Everyone gets accused of being in a cult now. EA, e/acc, woke people, anti-woke people, Trump supporters, people with ‘Trump Derangement Syndrome’, Corbyn supporters, Brexit supporters, FBPE. All cultists. I’m sceptical!.
"Not everyone who believes in AI risk is in a cult, but there is a cult involved here.". This seems to me a sober and true take. (And respect to @RokoMijic for agreeing it is "probably a somewhat accurate characterization.").
6
1
39
Has anyone actually written some long read about what we would expect to happen to the value of capital if the value of labour falls a lot? If I have £X in equities today and we get AGI that replaces lots of cognitive labour soon, how high does X have to be for me to be 'safe'?.
9
2
38
This piece I wrote on how to do well on Substack may be useful to some people who are participating in the @effective_ideas blog prize, as well as for other EAs trying to promote their content.
1
3
36
This idea was ahead of its time. 2024 is the year we finally build these things 🫡.
Serious question for #YIMBYTwitter: where height restrictions per se are the issue, why haven’t developers started building vertically but *down*, underground? . I mean windowless apartments don’t sound ideal, but probably less bad than unaffordability and homelessness
7
1
37
The fedora isn’t particularly bad, the reason Eliezer isn’t the right guy for mainstream interviews is the haughty and condescending attitude towards interviewers and the odd, confusing metaphors.
I distrust any EA/rat that says the fedora is bad for the movement. if you care more about optimizing vibes than the truth, you should retire from saving the world and take drugs / raise a family instead.
0
1
34
A classic recipe for a hit tweet is to reference something that seems like it might be obscure knowledge, but in fact lots and lots of people know about. Dunning-Kruger, the Monty Hall Problem, Coastline Paradox - all ripe for virality!.
Mathematically speaking, both countries have infinitely long coastlines so we can’t say which has a claim on the name.
3
1
35
I don't think it's good to claim that people who publicly denounce EA are risking unnecessary death/suffering. Or if we do claim this, it makes sense to also note that anyone who fails to donate all their 'spare' income to charity is also causing unnecessary death and suffering.
6
0
34
The only stock recommendation Samstack has ever made is to buy Meta (made in late January). Meta is now up ~125% since then, compared to 18% for the S&P 500. My blog may have the best stock picking record of any blog ever! Why not subscribe?
3
0
34
Meta is up 183% YTD, Asana is up 48%. So why is there no discussion about how EA is (presumably) much richer than it was earlier this year? Does Dustin have less money in those companies than I would've thought? What's going on? This seems great!.
7
1
31
I think it’s unwise to tell people ‘you are in a special intellectual environment that means you know the truth about the world before everyone else’ for a number of reasons.
being in tpot is literally one of the highest leverage moves you can make in today's world. just by being here, you are lightyears ahead of the general population and can see things MONTHS before they do -- a prescience which you can leverage in many valuable ways.
2
0
33
Imagine you're a star columnist at a major publication who is basically subsidising other writers at the publication. What's the main reason for staying rather than making £££ on Substack? Status? Enjoying the culture of [publication]? Altruism? Status quo bias? Something else?.
11
2
32
You can use Anthropic reasoning to update against insect/shrimp sentience btw. If insects/shrimp were sentient they’d outnumber humans by like 100 million to one. The fact you’re a human and not an insect is a huge update on insect sentience!.
11
0
32
@georgmolzer @rgblong @young_opsimath But presumably very few humans actually would deliberately antagonise a tiger, even if they would find it funny if they succeeded.
10
0
31
When I last posted about EA, some people said it was counterproductive to only talk about the things I dislike, so here's a defence of how weird you all are.
0
3
31
When you ask Deepseek-R1 about China, you can see its thoughts about how to answer your question while promoting the CCP party line.
2
2
30
Marcus may be wrong about a lot of things but it seems pretty reasonable to say that there is not yet a GPT-5 level model, and that many people expected that there would be one by now.
Never ask. A woman her age.A man is salary.Gary Marcus to define what he means by "GPT-5 level"
5
0
29
@cunha_tristan Probably one of the biggest mistakes by an EA Org in history alongside 80k formally recommending the 'Ponzi Scheme to Give' career path back in 2012.
1
0
27
I often disagree with him, but I do think that Cummings' discussion of how people react to big events is mostly spot on (although he should probably credit Tetlock here).
1
3
27
@effective_ideas This was inspired by @Noahpinion's excellent piece but aimed more at people who don't already have a platform and need to figure out how to even begin promoting their content.
0
1
27
No, but probably fairer than if humans were making the decisions.
2
2
25
Honestly I have always felt mildly embarrassed bringing up AI as a potential serious problem to friends/family. It would be good if there was some way around this!.
1
0
23
Correction:.
@sam_atis Note that these were not proposed messages. These were what they wanted the architecture and design to evoke without using language.
0
0
26
I guess I knew this was slightly rude/weird/EA-ish but was quite surprised at how extreme the reaction was. It was immediate and really much more angry than I would’ve expected. I have a newfound respect for @bryan_caplan.
1
0
26