I am naturally quite an underconfident and risk averse person, so whenever I though of doing anything scary, I would focus on the low probability of success and not do it. Upside Bargains is a framing that shifted my focus to what would happen if the risk did or didn't work out, because specific risks (Upside Bargains) have massive upsides and minimal downsides. So, even though they're unlikely to work out, I'm happy to take Upside Bargains/ risks, not because I think I'll succeed, but because I know I'll be fine if I don't.
So, in this post I'll go through (1) how to think about upside bargains, including concrete examples; (2) why I think we don't take them more often; and (3) provide strategies to overcome this, starting right now.
This definitely isn't my idea, but I think it's really simple and useful, and I'd like to share it as best as I can. This is based heavily off Neel Nanda's blog, Peter Wallich's workshop at EAGx Cambridge, and Annie Duke's Thinking in Bets.
Concept
'Upside Bargains' are events that have a <50% probability if upside, but whose expected upside is large and expected cost is small, (so they have positive EV).
- Examples
- Asking for advice or favours (e.g. cold emailing). Cold emailing is unlikely to succeed, but it can be really good if it does succeed, (e.g. you can get free advice/ connections or be more likely to get a job), and no-one cares if it fails.
- Hobbies (e.g. weird sport or blogging)
- Making new friends
Barriers
This section is to help you realise that our undervaluing of these bets comes from biases- heuristics that worked in the past but don't work now. (I'll also talk about solutions after).
Thinking in terms of most likely outcomes ('most likely fallacy') or what actually happened ('resulting').
When we take risks and it doesn’t work out, (e.g. putting a lot of effort into an application that gets rejected), we often think it was a bad idea. Motivated reasoning tells us that we didn't want to job anyway, and hindsight bias makes us think the rejection was inevitable (even though it could have easily gone the other way). Because of these biases, we think it was a bad decision, even though it may not be- as it may have positive EV.
The bias of resulting also goes the other way: we tend to think that people who risk everything and succeed against the odds made good decisions, but the EV of their decisions could be very low.
Present vs future you- (temporal/ hyperbolic discounting)
To take an upside bargain, you have to risk rejection so present you pays the cost, while only future you can reap the benefit. This asymmetry is well documented, (as people would rather take less money now than more money later, or present you doesn't want to go to the gym but future you wants to have gone to the gym) but irrational, as, in the future, you will become future you.
Possible solutions (more below) include:
- Automation: make the decision once to go to the gym at a certain time, and make it a habit. This is useful because we love the status quo. With upside bargains, make a decision to try a new sport/ society every week.
- Make the upside more concrete: Imagine, as vividly as possible, that you're future you and the bargain went well. (This should be so much more pleasant than the rejection is unpleasant).
Social rejection and loss aversion
Obviously, asking a stranger for a favour or applying to something really selective is scary as we fear rejection. This is because of evolution, (as, in the past, being socially rejected could get you killed), but this isn't the case now!
- Sometimes we can't even see non-standard, socially endorsed paths. (Even Neel Nanda (who came top in Cambridge maths), found it hard to take time out for AI safety research as it's not what he was 'meant' to do).
- Solution: Expand the range of options that we consider by brainstorming options under a time limit, and having a really low bar for what we write down.
- And loss aversion means we're more afraid of the concrete downside than losing out of the hypothetical upside
- But this is irrational: (people would rather keep an item than swap it for a nearly identical item).
- Solutions:
- Imagine the worst case scenario: does it matter a month, a year, 10 years from now?
- Get used to taking low level risks and build up. Rejection is inevitable, so it's worth getting used to.
Solutions
Ways to see the upside
- Make the upside more concrete: imagine, in as much detail as possible, the world in which you saw the upside. This is prehinsight/ backcasting.
- Redefine success:
What makes a great decision is not that it has a great outcome. A great decision is the result of a good process, and that process must include an attempt to accurately represent our own state of knowledge. That state of knowledge, in turn, is some variation of “I’m not sure”. (Duke, Thinking In Bets).
Therefore, most decisions with positive EV are successes- rejected applications are good decisions.
- Benefit of taking these risks: Taking Upside Bargains makes you more likely to try them in future, and this is beneficial overall (because they're positive EV). Rejection is inevitable, but taking small Upside Bargains makes you less likely to fear it.
- There are also specific benefits to trying stuff, e.g. applying to stuff is useful as it makes you collect evidence about what you're good at.
- Alternative framing: gained info- if I try something and it doesn't work out, I've learned something (at the very least I've learned what I'm not good at, or I've become more resilient).
- Opportunity framing: when asking someone for help, you're giving them an opportunity to help you (e.g. Neel Nanda finds it fun to give advice), so you're giving them free option value (as they can always say no).
- There might be a cost of saying no, but the busier and higher status they are, the more accustomed they are to saying no, so the lower the costs
Increase probability of success
- Premortems- imagine it going badly so you can figure out how to avoid this. You can also use pre-hindsight for this- in the world where it went well, what did you do?
- If you're looking for advice, note that some people (like Neel Nanda) have documents on how they like to give advice.
Reduce fear of downside
- Another framing: flip the roles of the people in the situation, e.g. if you're trying to reach out to a friend (who you haven't contacted in ages), how would you feel if they reached out to you?
- “What are the consequences of each of my options in ten minutes? Ten months? Ten years?” (Duke, Thinking in Bets)
Takeaways
- Be ambitious! Not because you think you'll succeed, but because you know you'll alright if you fail.
- Set a 5 min timer and brainstorm solutions to a problem. Have a low bar for writing stuff down!
- Try everything once e.g:
- productivity tricks
- Hobbies
- Friends/ dates (e.g. I asked someone out and it didn't work out but that's ok)
- Optimistic job applications
- Do the crazy ideas that pop into your head
- Post in the EA forum
- Caveat: Upside Bargains are defined as actions with low costs associated with them, so they don't include actions similar to Sam Bankman-Fried's.
I think there's a minor mistake. "as people would rather take more money now than less money later" I think this is flipped and supposed to be the other way around? ""as people would rather take less money now than more money later"
Fixed- thanks for flagging!