The bus ticket theory of genius

[jacob]

http://paulgraham.com/genius.html

One of the features of the web-of-goals is conditioning nodes and vectors (stocks and flows) on being useful. The bus ticket theory suggests that this is a bug.

[jacob]

Fair point that WOG doesn't create genius, but ... my point was that [WOG] might prevent genius events from happening by design.

[Mister Imperceptible]

Echoes of Taleb’s lottery tickets.

It feels intuitively true that by leaving nodes unconditioned that it remains possible for them to capture beautiful accidents.

I think there is a political corollary to this also, that more freedom/less centralized control is better because top-down command structures limit possibilities only the creative individuals and groups can realize.

Thanks for the introduction to Graham, more rabbit holes to go down.

[5ts]

I agree with Jacob. I can't imagine how generalization can create genius since obsession/specialization is a critical component of genius. It's also important to distinguish generalization from serial obsession/specialization. I can imagine how serial obsessions can create genius to a degree.

[George the original one]

I don't think generalization prevents genius. After all, genius can also be taking the disparate ideas a generalist is exposed to and applying them in areas where they haven't been applied, thus advancing in this area. An example is taking the perfume atomizer and applying it to internal combustion engines, thus inventing the carburetor.

[Jin+Guice]

I disagree. First off, the web of goals needs a meta-goal, which for our purposes is ERE. The meta-goal can be refocused on the obsession. I realize that at some point web of goals is about breadth and genius (according to this theory) is about depth, but "he who dies with the most skills wins" IS just a joke, right? In other words, web of goals doesn't specify exactly how many skills one needs or how deep one should go in any one area. I thought the main idea was that you could go shallow on some easy stuff and fairly deep in multiple areas. In my mind, it can be really deep in a few areas or pretty deep in multiple.

No one of the geniuses described in the article literally did one thing all the time. There will be other elements to the genius life. These elements need to be arranged in a way that supports the meta-goal. I might even argue that genius causes web of goals thinking because being obsessed with something will cause someone to line up their life in a way where everything supports spending time on the obsession.

The article discusses how Newton had 3 different obsessions, only one of which "paid off." That sounds like a web of goals lifestyle to me.

I don't think that, if we assume the bus ticket theory of genius, it's fair to say that it lines up perfectly with web of goals. There is a trade-off between breadth and depth. But I don't think that web of goals mandates that "you must go as wide as possible" and that the genius "must go as deep as possible."

[black_son_of_gray]

my point was that [WOG] might prevent genius events from happening by design.

I agree, with the tweak that it might prevent some genius events from happening, but not necessarily all.

Basically, the criteria of "usefulness" in a web-of-goals limits the total amount of serendipity that can occur, and genius events rely in part on serendipity. Serendipity itself has a usefulness* in a web-of-goals—I just have no idea how it can be accurately valued. [Another example of the "I optimize what I can measure and ignore what I can't" fallacy.] And yet, if a web-of-goals is very large and diverse, there would probably still seem to be opportunities for serendipity between the elements due to the sheer size of the thing. Every element can't possibly be directly connected to all other elements, so while it may be dramatically reduced, the possibility of serendipity hasn't been entirely eliminated. A small web, though...

*I guess another way of stating this is that a very tenuous web-of-goals, without a lot of strong connections (and perhaps goals that seem to be at odds with each other), should have a higher amount of serendipity/genius events—that is, previously unrealized or unusually powerful connections are revealed in the web. They ultimately become useful web elements, but only after the construction, not before/during construction.

In this sense, the "best" web-of-goals is the one with the most overall lifetime usefulness vs. the one with the most usefulness at construction. That web-of-goals likely has some slack to allow for serendipity vs. a tightly engineered one. How much slack, I don't know, but it's consistent with how I've been feeling recently, that in the end these systems need to tuned "by feel".

Of course, I don't see why a two-pronged strategy can't be done: have a tight core web-of-goals that takes care of the core essentials of life, then dabble in the fringe according to your interests in your spare time or odd hours here and there. It'll be 90% efficient instead of 97% efficient. But it invites in the possibility of life-changing upsides!

[classical_Liberal]

I disagree as well. This world already has too many technical, specialized geniuses. What we need are interdisciplinary collaborations to actually use the technical abilities towards real human progress. This requires generalist knowledge and thought processes. If it wasn't for Haley, how long would have it taken to figure out that Newton had already solved laws of motion? What good is a vaccination if we can't figure out how to get people to take it? What good is a technical solution to most of the worlds problems, like ERE, if we can't get more people to live the lifestyle?

[Toska2]

WOG

PRO : leaves freedom to develop a broad set of skills and disciplines that could be tied together

CON: some skills need machines and network only available in very specific fields to fully develop

Where one is at in the development of the field and the type of field determines the need, therefore could be either PRO or CON.

(Ie I cant find the NPR program which stated longer migration bird migrate by daylight and not temperature. Therefore they are less able to adapt as vegetation/food that appears earlier. This obsevation can be done with time and passion alone from1600-2100. Exoplanet observation is from 1990-?)

[fiby41]

More risk=more reward: I don't believe in that anymore. Only the MBA types purport that when what they really mean is you risk your capital for their sake. The rate of return after accounting for margin of error (me being wrong) should still be above SWR else I'll just leave the amount in a bank account.

Tieing nodes together: The aim is to increase the euclidean distance (d) between two adjacent nodes. Lower d means higher interference and overlap in the power spectral density of channel while higher d means a lesser number of nodes so less resilience.
Solution is two-fold, decrease the number of nodes or adjust the arrangement of nodes.
On one end is phase shift keying constellation diagram with only 2 nodes and d=pi for binary.
On the other end is quadrature amplitude modulation constellation diagram with nodes 2^n or greater for n>=3
So looking into this to increase the capacity of the channel is only beneficial if you have greater than n=3 areas of competence.

[7Wannabe5]

I think I missed the note on WOG being conditioned on “useful.”

The article linked ties in neatly with a book* I am reading on the topic of innovation. Apparently, top down R and D is not even efficient at the corporate level. Individual hobbyists who innovate purely for their own self-interest are more likely to create value than those who are guided by profit motive. IOW, something like “fun” is more motivating towards innovation than something like “money.”

*”Democratizing Innovation” by Hippel

[7Wannabe5]

It’s also true that those in the grip of an obsessive interest will naturally tend towards whittling their “needs” down to a minimum. For instance, during the era of high interest in natural history, all sorts of rackety adventures were undertaken in search of new species.

Also, in the realm of FIRE, it does seemlike those seeking “freedom to” rather than “freedom from” are more likely to quickly figure out how to reduce expenses. Of course, this tendency would also be well correlated with less robust and or more hare-brained solutions.

[Jin+Guice]

Rereading the OP I see that @jacob is highlighting the need for nodes to be "useful" as a bug and not the whole WOG approach. I agree that demanding something to be "useful" has the potential to inhibit the serendipity the article talks about. I always interpreted "useful" as "useful eventually, probably." Saying XYZ must be useful in W years sounds a lot like a tightly coupled goal. I assumed that "fun" was allowed to be something I consider in choosing my nodes. Once you're at the point where reducing expenses is monetarily inefficient* (i.e. I make more per YMOYL hour than it would cost me to pay someone to do this), I can see no reason to consider "usefulness" over "fun/ interesting," as long as it doesn't set me back.

*I'd want to be pursuing fun/ interesting before this point, but that's my preference.

My meta-goal for ERE is something like "freedom to fuck around" or "freedom to do whatever I want whenever I want," my point being I think the WOG approach allows for serendipity as long as you don't take the "usefulness" too literally.

[Tyler9000]

I don't think generalization prevents genius. After all, genius can also be taking the disparate ideas a generalist is exposed to and applying them in areas where they haven't been applied, thus advancing in this area. An example is taking the perfume atomizer and applying it to internal combustion engines, thus inventing the carburetor.

+1 Genius isn't about mastery. It's about inspired insights. And it's ok if one node in the web of goals is picked according to passion rather than utility as long as the system is robust.

In the design industry they talk about "T-shaped" designers as the ideal. That means they have very broad but shallow exposure to many different subjects, but one deep interest where they're a subject matter expert. The interest provides the drive and ability to execute, and the breadth provides the context to draw connections that spur innovation.

Re-read the section about the paths to new ideas in the context of a web of goals that makes those paths available. Obsession provides the drive, while the WOG provides the useful connections. One without the other isn't nearly as powerful as both in combination.

Another way to think of it is to simply appreciate the usefulness of pursuing a personal interest in your web of goals. Even if it is not inherently productive, the inherent passion can be a net positive not only for your happiness but also for maximizing your opportunity to stumble on something where you realize you can make a real difference.