Speculation Is (Not) Good
Good or bad? Who can tell really?
Mountains turn into clouds, fishes merge with water and become indistinguishable from their surroundings.
I think context is needed to make the argument on why or why not speculation is good.
Speculation can be dangerous, similar to making assumptions about reality from a third person perspective, then shifting into a first person perspective and getting punched in the face, so to speak, for mistaking the map for the territory.
Bad speculation
The really bad type of speculation arguably is speculating on things one knows nothing about with resources one can't afford to lose.
There's no net gain for the individual driven by ignorance gambling with necessity, although on the other hand, the collective gains information about what doesn't work, i.e the speculator becomes an inadvertent teacher through their losses for others to take note.
Good speculation (Individual Level)
A goodness of speculation is developing good judgment over time via calibrated feedback loops and being able to distinguish signal from noise.
This ability is quite useful across any domain requiring prediction under uncertainty such as in career moves or creative ventures, but it can only come from first having to make bets, be wrong, feel the consequences, and adjust.
Good speculation (Collective Level)
Another goodness of speculation which I'm gradually giving more awareness to is on a collective level and is more present in decentralized settings. It's the overall system gets better as resources flow away from poor speculators toward better ones. Speculators lose in proportion to how misaligned their models are with reality.
Broadly and in this sense, speculation is a positive sum game, because collective intelligence emerges from the aggregation of distributed bets. The system learns faster than any individual could.
Centralized vs. decentralized speculation
Speculation tends to be gatekept by a committee that decides what's worth betting on and which hypotheses get resources in centralized systems. Much reduction on variance and discovery happens here.
In decentralized systems, speculation is broadly democratized. Anyone can bet on their vision. Most fail, but the hits, when they come, tend to be genuinely novel because they weren't filtered through consensus thinking.
Individual vs. collective tradeoffs
For the individual, good speculation requires asymmetric payoffs: limited downside, unlimited upside. Bet what you can afford to lose on outcomes that could transform your position. Partly, antifragile speculation.
As mentioned above, even "bad" individual speculation can be valuable if losses are distributed and learnings are shared, given the collective subsidizes individual learning while capturing the aggregate wisdom.
Zero-sum vs. positive-sum framing
Financial markets are usually pictured as zero-sum (your gain is my loss), but even there, speculation can be positive-sum when it:
- Allocates capital more efficiently than central planning.
- Provides price discovery that informs real economic decisions.
Outside of finance, speculation on ideas, methods, relationships, life paths is clearly positive-sum in the sense that I'm speculating that learning skill xyz will pay off, I'm not taking from others and rather exploring a possibility of expanding my surface area of opportunities for myself and generating valuable information for those watching.
Point of the matter
The deepest insight is that speculation is unavoidable. I find it fascinating that even "not speculating" is a speculation that the status quo will persist or that passivity is safe.
Speculate consciously, with appropriate stakes, in domains where one can learn from feedback.
Which games to play, with what resources, over what timeframes?
This comes back to developing judgement again. Every individual speculation is always embedded in larger collective dynamics, so do well to also remember that one's edge comes from seeing both levels simultaneously.
Thanks for reading!! Share your thoughts below on the comments.
Posted Using INLEO