Dan Wang 2019 Letter

Dispersion is the process of a heuretic impacting the world

Is the rate of scientific progress slowing down?

The tech tree is a scaffolding inside the knowledge frontier

Technology has enabled progress to be a net positive

Singapore has reached the limit of Type II (catchup) progress  - the strategy that got it where it is today will not continue to move it forward

Tech startups can only exist in deindustrialized societies

# §Type I and Type II Progress
Imagine human capability as a high dimensional blob.[^1]  There are some people who live on the surface of the blob (let’s call it the [[Knowledge frontier]] ) - they have access to the best of the knowledge and resources humanity has to offer. *Pushing this frontier and expanding maximum possible capabilities is Type I Progress*.
There are many people who don’t live on the surface of the capability blob - most people who live in ‘developing’ nations, for example.  *Actions that expand these people’s capabilities towards the frontier are Type II Progress*
![](BearImages/1F49C0F6-549E-4B98-B0B1-0176C38B83DC-792-000235616E764616/6B79250F-C44C-4ABA-BE8E-F5F3CBD68654.png)
As William Gibson said [“The future is here — it’s just not very evenly distributed”]([William Gibson - Wikiquote](https://en.wikiquote.org/wiki/William_Gibson)) You could think of Type I progress as expanding what the future is, and Type II progress as distributing that future. 
There are even some people who might live a little beyond the surface in some dimensions. These are the innovators, the creators of new knowledge. 
This is a great example of Type I and II progress
![](BearImages/AAAAEAF2-C3E6-4198-BA89-A6257860C66B-75602-000348351FC9E35C/9D76763E-DA30-43B3-B336-8344A32A5A94.png)
Here’s a simplified version of the above chart.  You can see how from 1960->2006 Northeast Asia and North America were pushing the capability frontier forward (on the total possible output per Hectare and total output per worker respectively.)  China as a whole clearly made progress but stayed well within the capability frontier.
![](BearImages/DC29C3A0-429C-4AFD-A265-2B07D90B5A47-792-00023548B4ED6900/F5F33A50-D8F8-4FAB-B86B-1811A962C8BB.png)
In [[Lever of Riches]] Mokyr writes: `“At any moment there is a large gap between average and best practice technology; reducing this gap by disseminating the techniques used by producers at the cutting edge of knowledge is technological progress without invention.” `

Type II progress is this progress without invention. This distinction may seem trivial but [[Type I and Type II progress are often talked about in the same context]].  [[Explicitly calling out the distinction between Type I and Type II progress and thinking about them separately will lead to more action]]

[^1]: According to a mathematician friend, the correct term for this is a “compact” but for  the sake of evocative language I’m going to stick with “blob”

---
References:
[The Illustrated Guide to a Phd](https://www.notion.so/innovate/The-illustrated-guide-to-a-Ph-D-e4b0ed9ae1b74de9b8ee981f27db381d)
[Productivity Growth in Agriculture](https://www.notion.so/innovate/Productivity-Growth-in-Agriculture-An-International-Perspective-Google-Books-6f2195eda23d40f8ba21890fcc189201)
Images
![Original farm progress chart](https://i.imgur.com/7bwKDFY.png)
[Capability Blob](https://lh3.googleusercontent.com/v7YtSoUqTzwIAkFUQeN06AtVwqHazuySotlmrI1thsCMDx8bAh1dnL2-G2KPyQ15NQqNRG9s1yHnIXsoXb6NHgkrrKI9WKnJGCCbtmOoVvRi_7xIVYT7OOWl7tZDCJJx00OKE4F-DnNZhhzsxqmq2cZGoVIc_sLvSxXv6qUH3cJiELraDztvEY1gxmHTy_trBFOIxpS53GlEkAq257PuYYvFxGjiUtuQxQpj4U7gPo0zwWD8L64eE-dnQmaW_ZEgxMi5s45LjvDYeOz12gdYB8zJZEoZRkHOxIcHWBte55YWpXi_JXTJuEKeE427ZU-ZbPbvCzf7YTvfL2q6U4xiaC9ue9Z_ec0l0KoIlCNuOkhKAZZXTsmI6tMZ5AMeNFAjPUG8utDG5ZbjEhHUaudUGu_5lfbG2E8TJBWgNq72MP9M3lIP-leSkIkmCkzPazEcBDDoRi-0C8OlgsdlbPcWVArgtarrzYcWP14nWa9gP0oGKUmjClh1rGY51YI5FEhKNMrBmrlXLwS9X_yxFY3LfOCDvmXsIXIszm8OVETrIXtbiU-SqMAjgnr-PkUjfIGFIa3cX_tZSr5s2KkQqP5ZwOBnsVCpFKsXd2XzBq4kgQwNota4h4D4aSJNmaHatPhSGywGtDEEyEpCz0KhGiK7OrO1o8hLfuY2a6F5p53NhP0SNB1_lxl3CSFhu3Vmqam1ULTnrDemeO2So6Ri3aIULsTPgmeQkq9e0gG-S7Mc6migYdXSrg=w898-h947-no)
[Simplified Farm Output](https://lh3.googleusercontent.com/KnjVQiEJKUa5n-saje01_MSCME3XOgtKkt6pMV4E8C-pILQkh2axGI_Y7hO2gmL0PdVanhGsUX78qu0YJa9uZrfmhRwY1LrfZWdW4srlsn8LYciadYNfozpT-jviqkG7eJRWQY09h65xhv4T1TfYtdu_TuEHSic-5iA7DzhdwBcmpw89r-FSQ4Vv2BU5DGcsJlX9adCinYGF9r8rRjk0IV4lmQ-7303dI36qQqM8bCBWv2SYK55Y8HPczPt5dohXIcN0pUcjF8Tso9Kxr9W4ZsbILFmIHhrznUPESC51CfqeYNIOcvG_zdlLot1LQKNjgWiqGj7dg6sq0AHfkn0pzwmniu1JS2r5i4h6YtWFFI0d64d9aH5jPfdZbTlIDi3O3qY5s-b4wVAuZA5xPaPJ_P7N9a_JkR-7PH_vcecEN2_6BtFXVtp7XPVMbiPke1pVsVWqHVkxMgHDbABY8Vm7E-tM_Fwv7f7Ki6SOTY54H7H9c6LJQaHwEPHcIqxSwpmVvwMhXV4u492hVW0OzP25NLmi3zfUNSIBoz8MmrPej7_iQXvhT59-MVkNBQV7yD79xDldiZh2yLBP1gTyEtK723gLIjq9rjF9b_VK8JXKaLZO63hQJN4jMYZxWWjmZsFyZWq7chyuhfZtwTVRhrWR9az5noKVGkG8FP3ga5Pl28QWZitrjC4kAmITAvapE6usPTXUQXFOzKS_KyKDmqdn6-CkN-0gjIcpGjHyhZ2J-hgkROd27w=w1483-h804-no)
              ---
References:
[The Illustrated Guide to a Phd](https://www.notion.so/innovate/The-illustrated-guide-to-a-Ph-D-e4b0ed9ae1b74de9b8ee981f27db381d)
[Productivity Growth in Agriculture](https://www.notion.so/innovate/Productivity-Growth-in-Agriculture-An-International-Perspective-Google-Books-6f2195eda23d40f8ba21890fcc189201)
Images
![Original farm progress chart](https://i.imgur.com/7bwKDFY.png)
[Capability Blob](https://lh3.googleusercontent.com/v7YtSoUqTzwIAkFUQeN06AtVwqHazuySotlmrI1thsCMDx8bAh1dnL2-G2KPyQ15NQqNRG9s1yHnIXsoXb6NHgkrrKI9WKnJGCCbtmOoVvRi_7xIVYT7OOWl7tZDCJJx00OKE4F-DnNZhhzsxqmq2cZGoVIc_sLvSxXv6qUH3cJiELraDztvEY1gxmHTy_trBFOIxpS53GlEkAq257PuYYvFxGjiUtuQxQpj4U7gPo0zwWD8L64eE-dnQmaW_ZEgxMi5s45LjvDYeOz12gdYB8zJZEoZRkHOxIcHWBte55YWpXi_JXTJuEKeE427ZU-ZbPbvCzf7YTvfL2q6U4xiaC9ue9Z_ec0l0KoIlCNuOkhKAZZXTsmI6tMZ5AMeNFAjPUG8utDG5ZbjEhHUaudUGu_5lfbG2E8TJBWgNq72MP9M3lIP-leSkIkmCkzPazEcBDDoRi-0C8OlgsdlbPcWVArgtarrzYcWP14nWa9gP0oGKUmjClh1rGY51YI5FEhKNMrBmrlXLwS9X_yxFY3LfOCDvmXsIXIszm8OVETrIXtbiU-SqMAjgnr-PkUjfIGFIa3cX_tZSr5s2KkQqP5ZwOBnsVCpFKsXd2XzBq4kgQwNota4h4D4aSJNmaHatPhSGywGtDEEyEpCz0KhGiK7OrO1o8hLfuY2a6F5p53NhP0SNB1_lxl3CSFhu3Vmqam1ULTnrDemeO2So6Ri3aIULsTPgmeQkq9e0gG-S7Mc6migYdXSrg=w898-h947-no)
[Simplified Farm Output](https://lh3.googleusercontent.com/KnjVQiEJKUa5n-saje01_MSCME3XOgtKkt6pMV4E8C-pILQkh2axGI_Y7hO2gmL0PdVanhGsUX78qu0YJa9uZrfmhRwY1LrfZWdW4srlsn8LYciadYNfozpT-jviqkG7eJRWQY09h65xhv4T1TfYtdu_TuEHSic-5iA7DzhdwBcmpw89r-FSQ4Vv2BU5DGcsJlX9adCinYGF9r8rRjk0IV4lmQ-7303dI36qQqM8bCBWv2SYK55Y8HPczPt5dohXIcN0pUcjF8Tso9Kxr9W4ZsbILFmIHhrznUPESC51CfqeYNIOcvG_zdlLot1LQKNjgWiqGj7dg6sq0AHfkn0pzwmniu1JS2r5i4h6YtWFFI0d64d9aH5jPfdZbTlIDi3O3qY5s-b4wVAuZA5xPaPJ_P7N9a_JkR-7PH_vcecEN2_6BtFXVtp7XPVMbiPke1pVsVWqHVkxMgHDbABY8Vm7E-tM_Fwv7f7Ki6SOTY54H7H9c6LJQaHwEPHcIqxSwpmVvwMhXV4u492hVW0OzP25NLmi3zfUNSIBoz8MmrPej7_iQXvhT59-MVkNBQV7yD79xDldiZh2yLBP1gTyEtK723gLIjq9rjF9b_VK8JXKaLZO63hQJN4jMYZxWWjmZsFyZWq7chyuhfZtwTVRhrWR9az5noKVGkG8FP3ga5Pl28QWZitrjC4kAmITAvapE6usPTXUQXFOzKS_KyKDmqdn6-CkN-0gjIcpGjHyhZ2J-hgkROd27w=w1483-h804-no)
              

[Web URL for this note](http://notes.benjaminreinhardt.com/§Type_I_and_Type_II_Progress)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/§Type_I_and_Type_II_Progress)




§Tech Tree Model of Heuretics

It’s rational to try something that failed in the past as long as you explicitly call out why you will succeed when they failed

Impact in innovation is analogous to fitness in evolution

Each future node in the tech tree model has some function that maps effort in with a success probability distribution over time

# A heuretic’s impact is not endogenous to the technology but depends on the environment
In the same way that a species has fitness *with respect to the environment* an invention or discovery has impact *with respect to the environment.*  You can think of the environment as attenuating a heuretic impact.  [[There are many heuretic environments]]  -  so while a technology might be worthless in one environment, it could be incredibly impactful in another.

Heuretics can never exist *separate* from an environment, so it’s stupid to talk about “potential impact” of an idea. It is useful to think about “what would the impact look like in an optimal environment?” 

To formalize this, you could think of a heuretic as having some ‘feature  vector’ **F** that you take a dot product of with the environment **E** to get some ‘impact’ scalar.  This piece is both important to the [[Design for roadmapping system]] and tricky. One possibility is that [[You should be able to do a singular value decomposition to find the eigenvalues of a technology the things that matter]]. 

A technology’s impact can change with the environment much more quickly than an organism’s fitness because technology environments change much faster than normal ones. These rapidly changing environments are one argument for why [[Technodiversity is important]].

The environment includes (working list):
- The economy
- Culture
	- Heliocentrism didn’t have impact for a long time because it was heretical 
	- On the flipside, Darwinian theory became embedded in culture
- Legislation (Regulations) 
- Infrastructure for dispersing a thing
- Politics around an idea
- Perceived value

It is useful to think about other technologies as *not* being part of the environment so that you can distinguish what the main blocker to impact is.  

The delineation between a technology and the environmental attenuation of that technology’s impact roughly maps onto the questions “will it work?” And “should we do it.” 

Environmental exogeneity is also important because it allows you to distinguish when a heuretic would be low-impact but could lead to a high impact heuretic for a fixed environment. [[The tech tree model suggests that explicitly drawing out tech trees is important]].

There is something about “carrying capacity” 

### Related
- [[neumannOneProcess2020]]
- [[Knowledge frontier]]
- [[Frontier Technology]]
- [[§Tech Tree Model of Heuretics]]
- [[Evolutionarily Stable Strategy]]





[Web URL for this note](http://notes.benjaminreinhardt.com/A_heuretic’s_impact_is_not_endogenous_to_the_technology_but_depends_on_the_environment)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/A_heuretic’s_impact_is_not_endogenous_to_the_technology_but_depends_on_the_environment)




Activity Space

Ondrej Dyck conversation 20 Sep 2021

Astera Lab Meeting Notes

Activity design indicates how power and coordination are distributed

§Axes for Looking at Funding and Organizations

James Cooper conversation 7 Sep 2021

# A list of Knowledge generating activities
This list is meant to capture both the abstract and concrete “moves” that people do to expand the [[Knowledge frontier]]

- Knowledge compression. This of course, can fall into reductionist fallacies. Is this different than knowledge distillation? Maybe distillation has to do with practices or tacit knowledge. 
- Knowledge formalization. 
- Operationalizing the knowledge in a serious context of use and making small adjustments ([[Learning by doing]]) 
- Understanding the dynamics of the personalities nominally working in a similar area
- Figuring out medium-term exciting possibilities  
- Running interference for people doing really weird things
-  Scaling the production of a thing (Industrializing a thing)
	- Scaling the production of a thing by automating parts of the production process
	- Scaling the production of a thing by mechanizing parts of the production process
	- Scaling the production of a thing by parallelizing 
- Making a research artifact so it can exist outside the lab that built it. 
	- This can look like making code presentable or putting a made device in a box that lets it be transported and handled

[[Research organizations have core activities that act as discrete units]]

### Related
- [[Different axes of where research gets done in the US]]
- [[Research has many orthogonal and non orthogonal classification axes]]

[Web URL for this note](http://notes.benjaminreinhardt.com/A_list_of_Knowledge_generating_activities)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/A_list_of_Knowledge_generating_activities)






§Academia Constraints

Peer review has failed as an institution

Peer review doesn’t catch duplications

Human capital theory of stagnation

People are incentivized to work on nodes higher up the tech tree

It is easier to have a multipoint mutation in technology than in evolution

Peer review gives peers veto power over each other

Startups and Academia are bad at addressing structural constraints

# Academics are incentivized to do peer-reviewed publishable work
Peer review published papers both determine an academics livelihood and status. Therefore, they are incentivized to spend their time doing work they think can be published. 

On top of this [[Academic credit is a first-to-the post system]]

[[People and organizations are all playing some game that has different ways of gaining status and power]] In Academia there are a couple of long term games:
- Gain [[Kleos]] through contribution to the [[Knowledge frontier]]
- Just get to scratch your itch to investigate certain things
- ….

[[People create proxy games for long term or hard-to-measure games]] and in academia the entwined proxy games are:
- Peer reviewed Published papers
- Citations on those papers
- Grants (which are usually based on papers) 
- Awards (which are usually based on who knows you which comes from papers)

It’s a safer strategy to publish lots of papers than to try to write one super-cited paper because it’s very hard to know which papers will be highly cited. Additionally everybody is impressed when you say “I’ve published a bajillion papers” while only a subgroup might know your super-cited paper by title unless you’ve [[hit the top of the columnade]]. 

Tenure is based on the number of papers you’ve published and the amount of grant money you’ve brought in.  Without tenure, academia is extremely rough. People who don’t think they can get tenure leave universities. [[National labs and other government research centers are a research alternative to universities]].

[[The peer review and citation system incentivizes people to work on things that other people think is interesting]]

### Related
- [[§Academia Constraints]]



[Web URL for this note](http://notes.benjaminreinhardt.com/Academics_are_incentivized_to_do_peer-reviewed_publishable_work)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Academics_are_incentivized_to_do_peer-reviewed_publishable_work)




Heat cycles for democratization and cutting edge experiments innovation and discovery

Heilmeier — Licklider — Braben — Elon quadrants of Research Management

Overlapping institutional constraints rule out several classes of creative work

Institutions either make things more legible or act as trust brokers

The knowledge frontier is a high-dimensional garden

The linear innovation pipeline is more harmful than useful

Grants only go so far

Research organizations have core activities that act as discrete units

Does the nature of the activities to create new technology and science change over time?

New kinds of ARPAs need new kinds of performers

There is a tradeoff between an organizational culture that is good at addressing market uncertainty and one that is good at addressing technological uncertainty

A taxonomy of research organizations and activities

Cross-disciplinary real research conferences are underdone

Boris Hanin conversation 30 Aug 2021

Venture Research, Bell Labs, and DARPA form non-orthogonal basis vectors over the space of innovation organizations

Precommercial scitech is a nebulous term for a certain class of frontier activities

There is no legible pathway through precommercial scitech

3 Sep 2021

2 Sep 2021

Glossary

New ways of achieving collective action enable new activities

Journalistic research is the analysis half of novel research

Spending more money on the same activity has diminishing or negative marginal returns unless it explicitly scales

Vision-goal spectrum

# Activity Space
![](../BearImages/E75C5294-7BFA-4584-AA70-A0CBFE31A1B5-2972-0005F6028EB76D67/E6EFAF86-53BA-4CC7-8A22-671DC2597375.png)

Activity space is the set of all activities that people can do that contribute to new knowledge and abilities.[^1] “Ideas” jump from activity to activity, getting added to or changed in the process. These memetic footballs merge and split, sometimes carried by one person between activities and sometimes being handed off to someone else. Obviously what entails an “idea” is obscenely [[Nebulous]] and there is no right way to draw lines between them, but I think the concept of activity space is useful, nevertheless. 

Ideas will inevitably become ill-shaped for a particular class of activity, for example when an idea that was previously well-served by curiosity-driven probing needs goal-driven systems-level work to get to the next level. There are many frameworks to describe these transitions: [[Technology readiness level - TRL]], the Basic/Applied research distinction, etc ([[Research has many orthogonal and non orthogonal classification axes]]). Of course, knowing what activities an idea “wants” at any given time is incredibly hard and there is probably   

Take an extremely ham-fisted account of modern semiconductors: the work to purify germanium and create diodes for radar (clear goal, product driven, iterative, experimental, timeline, large teams and parallel efforts) led to work to create germanium triodes (small team (3), speculative, experimental, goal but not use-case driven) which demonstrated some odd phenomenon that needed to be explained theoretically (blackboard, theoretical, white space). This theory enabled the group to hypothesize that the germanium could be used to build an amplifier (imaginative, playful, speculative). However physically realizing this idea was very hard, potentially impossible. The work to physically realize the triode involved a lot of just trying shit out (fat process, playful, but goal oriented). The team originally tried to build a FET but it proved too hard, but analyzing the failure (slack to try many shots on goal, failure analysis)

The list of all these knowledge-generating activities is probably too long to enumerate (it may just be a list of all types of human activity) but here’s an attempt [[A list of Knowledge generating activities]].

Not all people, organizations, and institutions are equally good at enabling different activities. Different personal attributes, organizational structures, and institutional incentives lend themselves to certain activities and suppress others. [[Overlapping institutional constraints rule out several classes of creative work]]. As a result, it’s rare for an idea to achieve its full potential within a single person/org/institution and no person/org/institution should be applied to all ideas. 

### Related
- [[Research organizations have core activities that act as discrete units]]

[^1]: You could see these as all the different things that people do to cultivate the [[Knowledge frontier]]. If [[The tech tree is a scaffolding inside the knowledge frontier]], you could frame activities as the inputs to creating new heuretics. (See [[§Tech Tree Model of Heuretics]]). 

[Web URL for this note](http://notes.benjaminreinhardt.com/Activity_Space)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Activity_Space)






Knowledge frontier

Hall—Héroult process

# Bayer Process
The process of turning [[Bauxite]] into [[Aluminum oxide]]

Invented:[[1888]] by [[Carl Bayer]]

Replaced [[Le Châtlier process for alumina recovery]]

### Related
- [[Bessemer Process]]

[Web URL for this note](http://notes.benjaminreinhardt.com/Bayer_Process)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Bayer_Process)






Little h hard problems need a bunch of engineering effort but no breakthroughs

Science+Nature Risks occur when you need to discover a new principle to relax a constraint

DARPA is more ideas limited than money limited

Gaurav Venkataraman conversation 29 Jul 2021

An indefinite mindset means that you are out of good ideas

Magic Leap is a good case study of a potentially amazing technology killed by startup constraints

People time is important when there is a clear critical path

# Big H hard problems are those that need a discovery to solve
Because big H problems need a discovery, throwing more money at them won’t necessarily make them go faster. In this sense, they can “fail” because you can pour infinite money into them. 

This is probably another way of seeing science/nature risks. [[Science and knowledge of the universe as raw resources/Nature Risks]]

Some people would call this research, but research feels more like a spectrum between Big H and little h problems. 

Another framing on this is that Big H hard problems are situations where there is a lot of uncertainty around [[Critical path]] and the thing that matters is clock time. [[People time vs clock time]]

Contrast to [[Little h hard problems need a bunch of engineering effort but no breakthroughs]]

Is Fusion actually a Big H problem? [[deanFusionPowerMagnetic1976]] 

### Related
- [[Knowledge frontier]]
- [[The linear innovation pipeline]]
- [[Quadrant Model of Scientific Research]]



[Web URL for this note](http://notes.benjaminreinhardt.com/Big_H_hard_problems_are_those_that_need_a_discovery_to_solve)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Big_H_hard_problems_are_those_that_need_a_discovery_to_solve)




PARPA Partners

Reinventing Discovery

Notes on making ambitious awesome things 15 Aug 2020

Progress Studies Summit

# Bridge Model of Knowledge
You could think of human knowledge is like a bridge. The extent of the [[Knowledge frontier]]  is like the depth of the water. 

When we collectively knew little about the world, the pylons were quite shallow and one person had enough bandwidth to go deep on multiple pieces. 

Now we have pushed out into much deeper water.  We need literally layers of specialists just for people on the roadway to touch the bottom. In this situation the connectivity of the whole thing

In this framework it makes sense that you need many more people to keep moving forward at the same rate. [[Low Hanging fruit theory of stagnation]]. 



[Web URL for this note](http://notes.benjaminreinhardt.com/Bridge_Model_of_Knowledge)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Bridge_Model_of_Knowledge)




A technology’s value is relative to the amount of value it unlocks

A heuretic’s impact is not endogenous to the technology but depends on the environment

It’s hard to discover where the frontier is

# Frontier Technology
A better term for “deep tech”  because it is vaguely falsifiable. Think about technology that are on the [[Knowledge frontier]]

Figuring out whether something is frontier technology is straightforward: “does this technology do something that no other technology could do before?”

Because  [[Describing an effect in an inventive manner is important for separating invention and discovery]] , ‘doing something no other technology could do before’ could entail two different things corresponding to the different parts of inventive descriptions. ([[Effect A can be used for purpose B entails describing the effect in an inventive manner]] Either it uses an old effect for a new purpose or a new effect for an old purpose. 



[Web URL for this note](http://notes.benjaminreinhardt.com/Frontier_Technology)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Frontier_Technology)




§Program Design

Most roadmaps suck

Good Simulations could be the ‘why now’ of roadmapping tools

# Good roadmaps work backwards from a goal
Working backwards from a goal prevents ideas from being locked-in by how people think about the problem right now. 

In a way, thinking backwards from a goal is what people mean when they talk about first-principles thinking. The usual interpretation is to start with physics but I would assert that while tying back to physical laws is important to ‘first principles’ thinking, it’s actually the end. Or perhaps ending each step with checking “does this violate physics?” Is an important part of the process.  Working backwards from a goal is ‘first principles’ thinking in the Aristotelian sense of not working off of the work of others, which *is* critical to working backwards. 

The working backwards approach contrasts with the map-and-then-roadmap approach, where you would survey the entire landscape and *then* create a path through it. [[Roadmaps have two pieces - the map and the path]] . In reality, the approach is probably a deeply impure mixture of the two approaches - jumping back and forth between working backwards and describing the present state of the world. [[Understanding a thing has to do with intuiting the affordances of a thing]] and in order to understand the affordances of the area you’re playing in you need to at least engage with the way it does work now.  

In a way there is a tension between working forward and working backwards. If you purely work backwards you ignore the accumulated wisdom of one or more disciplines, which is considerable if those disciplines are valuable in the first place. [[We stand on the shoulders of too many giants to give them all credit]]. And to compound the problem, often that wisdom (especially wisdom on the [[Knowledge frontier]]) is not well modularized and composable, so it’s hard to know what’s important and what’s ignorable without knowing it. At the same time, starting with “the state of a field” can be a trap in a couple of ways.  [[The tech tree is fractal]] so you could actually spend an infinite amount of time digging into exactly what the frontier looks like. Additionally if you spend a lot of time immersed in current paradigms you’ll start to see the world through those paradigms, which is problematic for going after goals that almost by definition have not been reached by the people who work in those paradigms. 

This tension between working backwards and forwards is embodied any time someone approaches a problem outside of their ‘area of expertise.’ It is perhaps best illustrated by physics. Physicists are constantly bombarded by theories from quacks who think that they have found a new theory of <life, the universe, and/or everything.> At the same time, people who in the past *have* found new theories have on occasion come from non-traditional backgrounds and *were* regarded as quacks, until they weren’t. 

Perhaps the correct way to think about roadmapping is actually similar to a backpropagation algorithm, where you do many passes going both backwards and forwards. You

Not working backwards from a clear goal is one of the ways that [[Most roadmaps suck]].  A big reason they don’t do this is that working backwards is *really hard.* It forces you to consider important questions about something that doesn’t actually exist. Asking questions about “made up things” isn’t a mode of thinking we’re used to. 

[[The Architected Innov􏰂ation Institute - Engineering Scientific Revolutions]] asserts that working backwards from a goal helps you evaluate possible paths to that goal with an unbiased eye. 

[[Treating Science Fiction as case studies from a future history]] might be one way to make considering important questions about things that don’t exist. 

Thinking about hard-to-think-about questions regarding things that don’t exist seems like an actual place for tools for thought. [[You could build tools to make program design more efficacious]]. 

Another framing besides ‘goals’ is answers to the question [[What does ‘working’ mean?]] for this particular program?

One thought is that perhaps you [[Work backwards with questions and work forward with answers]]?

### Related
- [[A lot of research feels like it is building lego pieces without any idea how they’ll be used]]
- [[Roadmapping could be thought of as backwards-chaining together a series of conception-only heuretics]]
- [[Technological roadmapping can give an idea of where technological evolution can go and how to get there]]
- [[Luke Durant Email 18 Sep 2020]]



[Web URL for this note](http://notes.benjaminreinhardt.com/Good_roadmaps_work_backwards_from_a_goal)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Good_roadmaps_work_backwards_from_a_goal)




# Hall—Héroult process
Process of electrolyzing aluminum oxide into aluminum

[Hall–Héroult process - Wikipedia](https://en.wikipedia.org/wiki/Hall%E2%80%93H%C3%A9roult_process)

Invented:[[1886]] by [[Charles Hall]] (HE WAS 22??)
Invented:[[1886]] by [[Paul Héroult]] (HE WAS ALSO 22)
[[1888]] [[Charles Hall]] opened production plant

Required:[[Electric Power]]

### Related
- [[Bayer Process]]

[Web URL for this note](http://notes.benjaminreinhardt.com/Hall—Héroult_process)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Hall—Héroult_process)






Research has many orthogonal and non orthogonal classification axes

The ability to say ‘huh that’s funny’ is important for discovering new phenomena

Technological and scientific paradigms are each based around exploiting a specific phenomenon

Theory can suggest where to dig for new phenomena

drexlerRadicalAbundanceHow2013

The Most Famous Loop - Article

There are four different levers of innovation

Podcast List

Kuhnsian knowledge creation and Popperian knowledge creation assign different first steps to the same loop

There are four different players in an innovation ecosystem

Tool accessibility is underrated for heuretic development

Anatomy of Next

# Heat cycles for democratization and cutting edge experiments innovation and discovery
![](BearImages/E919E8F4-5085-49E7-8D0B-0183A660047D-453-00019EDE6D9282A3/E7045B1E-F12A-4DA8-A62D-4E2C9445B30E.png)

Carnot cycles work by iso-entropic temperature increase (putting work into the system) then iso-thermal entropy increase (increasing disorder at high potential) then iso-entropic temperature decrease (getting work out of the system) then iso-thermal entropic decrease (increasing order in the system)   ([[The Most Famous Loop - Article]]) 

Also related to [[Conjugate Variables]] [Conjugate variables (thermodynamics) - Wikipedia](https://en.wikipedia.org/wiki/Conjugate_variables_(thermodynamics))

What can you abstract volume and pressure into?

PV = nRT
W = (1-Tc/Th)Q

Variable you have to play with:
- Temperature (potential) 
	- Theory vs practice  - this makes sense because 
- Entropy (disorder) 
	- Democratization vs. institutionalization
	- Big coordinated teams vs small uncoordinated teams teams
- Volume 
	- Tinkering vs Theory-driven building
- Pressure
	- Competition/Time scales (or 1/[[Slack - concept]] in system) 
- Work 
	- Progress 

Things that I would like to try to throw at this framework
- Democratization vs. institutionalization/bottom up vs top down/centralization vs decentralization
- Theory vs practice 
- Cheap vs. expensive equipment 
- Big vs small teams 
- Tinkering vs Theory-driven building
	- Theory driven building works well when you have a very complex system. Tinkering tends to happen when you democratize the system. System complexity increases as you exploit the paradigm. 
- Discovery vs. Invention 
- Competition/Time scales
- Type I vs Type II progress
	- This generally corresponds to whether you’re making new things or making existing things better
- Exploration vs Exploitation
- Formalization vs. informalization
- Revolutionary vs evolutionary work
	- Ideally when you run out of theory (are these the fruit on an S curve?) 

Loops could have infinite dimensions - Carnot cycle technically has four - but it doesn’t need to. Do you need to have an even number of dimensions? Yes because you need to have conjugate variables 

Are big and small teams isomorphic to centralization vs decentralization? On the one hand you could have a bunch of individuals working under the same forcing function - kind of like the NSF: is that decentralization or decentralization. Maybe what I mean is more small uncoordinated teams or big coordinated teams. That is isomorphic to centralization vs decentralization. 

You can also abstract further and think about any time you have these dichotomies where there are examples of dichotomies where the answer is “sometimes you do one and sometimes you do another.” 

Thinking about loops is important because you can shift the discussion from “how do we keep trying to push this one leg of the cycle as far as we can?”  To “when is it best to make a right turn?” 

Theory-driven observations happen in a low-democracy state ([[Theory can suggest where to dig for new phenomena]])  because they’re most effective when there is a small group of people who all know about the same thing. 

High-democracy observations can lead to more theory/new paradigms because they allow people to discover things that weren’t even theorized.  In order to create high democracy observations, observation tools need to become cheap and plentiful enough. Observation tools become cheap and plentiful when there is a lot of demand and use of the tools, which forces learning by doing and miniaturization. 

You can discover new [[phenomena]] either by p

### Related 
- [[Activity Space]]
- [[Knowledge frontier]]
- [[When should we defer to experts?]]

### References
- Rethinking Science with [[Josiah Zayner]] — see [[Podcast List]]
- [[Epistemic Trespassing]]



[Web URL for this note](http://notes.benjaminreinhardt.com/Heat_cycles_for_democratization_and_cutting_edge_experiments_innovation_and_discovery)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Heat_cycles_for_democratization_and_cutting_edge_experiments_innovation_and_discovery)




Any conversation can be a conference

Conversation with Alex Ray 25 April 2021

# Heretical scientists and cryptoacademics are natural allies
[[Heretical scientists]] and [[Cryptoacademics]] seem like two sides of the same coin. The former remain within the system but rebel against it and the latter still hew to the system while officially not being part of it. Both care about research and pushing the [[Knowledge frontier]] more than the academic system that is nominally supposed to be the framework for those activities. 

My hunch is that there is much less crossover between the two groups than would be fruitful. The actionable thing would be to identify people in both groups (which will be hard because both naturally hide to protect themselves) and bring them together. 

### Related
- [[Cross-disciplinary real research conferences are underdone]]
- [[It is feasible to consciously cultivate a scene of frontier openers]]

[Web URL for this note](http://notes.benjaminreinhardt.com/Heretical_scientists_and_cryptoacademics_are_natural_allies)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Heretical_scientists_and_cryptoacademics_are_natural_allies)






# Innovation & Scientific Progress - Paper
# Innovation - Scientific Progress - Paper
P2
- Many different people have asserted that economic progress slowed down
P3 
- Some metrics indicate science is slowing down and some indicate speeding up
P4
- contrasts Leonardo’s helicopter (high profile, low impact) with moulderboard plow (low profile, high impact)
- Asserts that you can tell when new technology is propagated by looking at productivity measures
P5
- Asserts GDP can be used as a measure of productivity
P9
- One possibility is that slowing down [[General purpose technologies]] have led to a progress slowdown
- One argument against [[General purpose technologies]] theory is that things unrelated to fossil fuels became better in the industrial revolution
P11 
- Catch up growth has become easier
- Global adoption of technology has sped up - see gap with spindle (130 years) railway (69years) 
- Calls all human knowledge science
- Calls it uniquee science when an asian country fits a prexisting technology to their needs
P12
- zero to one Peter Thiel claims the world is going through a period of spreading around innovations to get richer
- The question of why Chinese ‘catch up growth’ went from nothing to extremely fast after 1979 is important
- Because the Chinese catch up growth doesn’t follow the [[Solow model - ToA]] or other models seen before it isn’t just “catch up growth”
- It might be good to increase progress by giving up some scientific progress
P13
- population growth even without living quality increase counts as progress
P14
- If growth is faster than the populations reproduction rate it goes Into increaseing real wages
- GDP is actually a poor measure of scientific progress illustrated by the case study of Norway’s oil wealthghfrtewqs
P15
- Total factor productivity is unexplained economic growth beyond land, labor, and capital
- TFP might be the best measure of science growth of narrow economic measures
- One problem with TFP is that it takes variable amounts of time for ideas to diffuse
![](BearImages/468C15F3-5263-410B-AD7C-E1A903A49610-4098-00004DB4B00FB089/3C4AA20B-E558-4698-891E-1973EE331BC7.png)
P18
- One problem with TFP is that some scientific advances work *through* the explained factors - land, labor or capital
	- Example: drugs
- TFP assumes perfectly competitive firms at the knowledge frontier
- Ambiguity - is a management consultant improving worker productivity? It’s a new idea, but it could be seen as catch-up growth in labor
- Ambiguity around what is “on the frontier”
	- If frontier is what the status quo before the application of some knowledge, even applications of existing knowledge count as TFP
	- Other definitions of frontier lead to other TFPs
P20
- Things that increase welfare but don’t capture spending like FB google etc. don’t get captured in TFP
- Three points against above point
	1. You can make money on the internet
	2. People pay for things on the internet indirectly through device purchases
	3. Cheaper things on the internet shouldn’t effect GDP
- Claims no clear evidence that internet brings higher consumer surplus than innovations in the past
- Elasticity of demand for internet would indicate that it is not seen as a basic necessity
- Syverson shows that places with much smaller tech sectors have similar productivity slowdowns which argues against uncaptured gains because of the nature of the internet
P21
- Tech sector in US is smaller than people think
- Argument that “tech is in everything” doesn’t work either
- Productivity slowdown dates to 1973
P22 
- Number of researchers doubling every 16 years - faster than population growth rate
- Number of papers doubling every 9 years
P23
- Papers and scientists are inputs
- Patents invented in venice in 1474?
- Patents are a possible input to track inventions that are original, new, and significant
- Patents in Japan+Europe+USA are called triadic 
P24
- “Breakthrough patents” - patents that predict future citations of the patent [^1] have fallen
- Patent quality could be tied to citations and the $$ value of the patent
- Isaac Asimov listed 1510 great scientists [^2]
P25
![](BearImages/A5F5CDC5-8DA0-4D20-B17D-46A7EC500545-4098-00004FD9AE71EFA1/E8AC75B8-443C-40D4-B31D-59F4B2C8D183.png)
P28
- Table of different innovator lists
- Bunch and Hellmans have a list of 8583 advances[^3]
P29
- These measures show lowering level of innovation *per capita*
P30
- *Are ideas getting harder to find* is strong evidence for science slowdown
- Some pushbacks against ^^
	- Researcher pay is not going down*
	- Many “researchers” might not be doing actual research (basically just dicking around)
	- Science might not be as much a public good as I the past and requires more work to apply it
	- Researcher productivity decrease may be tied to business productivity decrease
	- The super effective research might be in areas where we’re not measuring
- Increasing age at first innovation seems to be corroborating evidence to bloom at al
P33 
- Rate of increase of crop yields per researcher has gone down
- Rate of increase of life expectancy has gone down
P38
- Counterintuitively, people have been trying to increase computation rates for a long time
- Moore’s law for chip size is slowing down
P40
- We may be living in a world where org/institutional improvements are easier than ‘pure science’
- Might be easier to “build out aggregates” than “boosting producticvity of component parts”
https://docs.google.com/document/d/1cEBsj18Y4NnVx5Qdu43cKEHMaVBODTTyfHBa8GIRSec/edit
￼

- What is the Fulton street society 
- What does it mean for science to be slowing down
- Need to reference to [[Knowledge frontier]]
- Isn’t average researcher pay going down if you count postdocs?
[^1]:See Kelly, Bryan, Dimitris Papanikolaou, Amit Sru,and Matt Taddy (2018) “Measuring Technological Innovation Over the Long Run,” National Bureau of Economic Research Working Paper, in particular pp.26, 46, 50.
[^2]: *Biographical Encyclopaedia of Science and Technology*
[^3]:See Bunch, Bryan H., and Alexander Hellemans. *The history of science and technology: a browser’s guide to the great discoveries, inventions, and the people who made them, from the dawn of time to today*
 
 - What is the Fulton street society 
- What does it mean for science to be slowing down
- Need to reference to [[Knowledge frontier]]
- Isn’t average researcher pay going down if you count postdocs?
[^1]:See Kelly, Bryan, Dimitris Papanikolaou, Amit Sru,and Matt Taddy (2018) “Measuring Technological Innovation Over the Long Run,” National Bureau of Economic Research Working Paper, in particular pp.26, 46, 50.
[^2]: *Biographical Encyclopaedia of Science and Technology*
[^3]:See Bunch, Bryan H., and Alexander Hellemans. *The history of science and technology: a browser’s guide to the great discoveries, inventions, and the people who made them, from the dawn of time to today*

[Web URL for this note](http://notes.benjaminreinhardt.com/Innovation_-_Scientific_Progress_-_Paper)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Innovation_-_Scientific_Progress_-_Paper)




Stripe press could provide low touch support to a community of frontier openers

§Type I and Type II Progress

Frontiers lead to absolute games instead of relative games

Frontier Technology

Big H hard problems are those that need a discovery to solve

Bridge Model of Knowledge

The Burden of Knowledge and the ‘Death of the Renaissance Man’ Is Innovation Getting Harder?

Possible FRO Structures

A list of Knowledge generating activities

Good roadmaps work backwards from a goal

Academics are incentivized to do peer-reviewed publishable work

It is feasible to consciously cultivate a scene of frontier openers

Many approaches to augmented knowledge generation ignore humans

Uptimal solutions in science and technology are rarer than you think

Type I and Type II Progress - publishing version

Heretical scientists and cryptoacademics are natural allies

Innovation & Scientific Progress - Paper

Production possibility frontier

The work to make the knowledge frontier useful does not itself always push the knowledge frontier

Understanding the US National Innovation System - Paper

# Knowledge frontier
Imagine human capability as a high dimensional blob.[^1] The knowledge frontier is the surface of this blob.  

I imagine the edge of human knowledge and capability (the [[Knowledge frontier]]) as an ever-expanding surface — everything from theories about the cosmic microwave background and the beginning of the universe to the ability to turn bauxite into aluminum via the  [[Bayer Process]] and the [[Hall—Héroult process]]. 

[[Frontier Technology]] is technology that sits on the surface of the knowledge frontier. 

The Knowledge Frontier is agnostic of the environment and has nothing to do with impact. 

* [[Frontier-dwellers tend to live in developed places]]
* [[Progress frontiers are geographically fractal]]
* [[You can only push the frontier on dimensions where you are at the frontier]]
* [[It’s hard to discover where the frontier is]]

It might be that the [[The tech tree is a scaffolding inside the knowledge frontier]]!

[^1]: According to a mathematician friend, the correct term for this is a “compact” but for  the sake of evocative language I’m going to stick with “blob”


[Web URL for this note](http://notes.benjaminreinhardt.com/Knowledge_frontier)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Knowledge_frontier)




Heuretics come out of people’s brains (TECHNOLOGY IS PEOPLE)

# Many approaches to augmented knowledge generation ignore humans
It’s tempting to see knowledge as the corpus of written documents and connections between them. These connections can be either explicit through citations and other references, or implicit through shared authorship, topics, or language. 

The theoretical backing for this idea is laid out in [[Undiscovered Public Knowledge (Paper)]]. The argument in UPK is that knowledge in people’s heads is subjective knowledge, which is subsidiary to objective knowledge which includes everything written down and the *implied* connections between those things. Thus, the goal is to more effectively retrieve things from the world of objective knowledge and draw connections between them. 

In other words, [[UPKD assumes that all valuable knowledge is both written down somewhere and interpretable when you find it]]. This ignores the prevalence of [[Tacit Knowledge]] in the world, especially at the [[Knowledge frontier]]. In other words, it ignores the the knowledge that is in human brains. This is problematic considering [[Heuretics come out of people’s brains (TECHNOLOGY IS PEOPLE)]].

I would argue that [[Most human knowledge is not encoded]], and may be impossible to encode without some fundamental discoveries. [[In order for knowledge to be discoverable and interpretable that means that it needs to be encodable]]. Even ignoring unencodable knowledge and accepting UPK’s knowledge model, most interesting areas are still replete with people still in the process of generating subjective knowledge and figuring out how to encode it (or who are too busy or are missing the skills necessary to encode it well.) 

Perhaps there’s a way around these two barriers, but I don’t see it. Instead,  [[Treat people as nodes and artifacts as edges]]

### Related
- [[Fast moving fields have lots of unconnected nodes]]
- [[Researchers are more important than the research]]
- [[Process Knowledge]]



[Web URL for this note](http://notes.benjaminreinhardt.com/Many_approaches_to_augmented_knowledge_generation_ignore_humans)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Many_approaches_to_augmented_knowledge_generation_ignore_humans)




# Production possibility frontier
[Production–possibility frontier - Wikipedia](https://en.wikipedia.org/wiki/Production%E2%80%93possibility_frontier)

Tradeoff between producing two goods, more generalized it’s like the [[Knowledge frontier]]

Economists focus a lot on the derivatives of the curves when maybe they should figure out how to push it out



[Web URL for this note](http://notes.benjaminreinhardt.com/Production_possibility_frontier)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Production_possibility_frontier)




neumannOneProcess2020

Every time you zoom in on a fractal, it presents the same pattern on a different scale

“Should we focus on Type I or Type II progress?” is the wrong question

Debating the abstract relative merits of Type I and Type II progress is a waste of time

Type I and Type II progress are often talked about in the same context

Being at the frontier in many dimensions helps advance the frontier

Alexey Criticisms on Type I and Type II progress

1 with developed and developing nations

# Progress frontiers are geographically fractal
[[Every time you zoom in on a fractal, it presents the same pattern on a different scale]]. This scale dependence makes it impossible to clearly separate between places that would benefit from Type I vs Type II progress.
In the United States you would expect California to be right on the knowledge frontier - it’s the wealthiest state with the most progressive policies. Yet there are still tons of farm workers who live lives comparable to residents of developing nations - they clearly are not at the frontier and would benefit more from Type II progress than Type I progress.
Zoom in farther to the San Francisco Bay Area. In only a few square miles there are people pushing the limits of quantum technology and infrastructure and policies that look like something out of a developing nation.[^1]
You can’t even say that individuals are the quanta of where type I vs. type II progress happens. Many researchers on the cutting edge of science follow dietary guidelines from the 1970’s. 
This inability to cleanly separate the two types of progress geographically adds evidence to the fact that  [[“Should we focus on Type I or Type II progress?” is the wrong question]].
Another consequence of this fractal nature of progress is that [[Type I and Type II progress don’t correspond 1:1 with developed and developing nations]].

In other words they are [[Nebulous]]

[^1] Some time-dependent evidence of this are the PG&E blackouts in October 2019. 
---



[Web URL for this note](http://notes.benjaminreinhardt.com/Progress_frontiers_are_geographically_fractal)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Progress_frontiers_are_geographically_fractal)




Specialization in research and careers in general has cut off a lot of possible innovations

Are ideas getting harder to find because of the burden of knowledge

# The Burden of Knowledge and the ‘Death of the Renaissance Man’ Is Innovation Getting Harder?
# The Burden of Knowledge and the ‘Death of the Renaissance Man’ Is Innovation Getting Harder
Interesting concept of modeling knowledge as a ring - similar to [[Knowledge frontier]]

Basically argues that the amount of knowledge someone needs to get in order to get to the frontier of a field necessitates more specialization which then requires larger teams to get the same coverage and you need larger number of people to do the same inventions.  Draws a clear distinction between this hypothesis and the [[Low Hanging fruit theory of stagnation]]

Very similar to the specialization theory in [[odlyzkoDeclineUnfetteredResearch1995]]: [[Specialization in research and careers in general has cut off a lot of possible innovations]]
### Published
2005
## Questions
## Thoughts
## Related
## Authors
[[Benjamin Jones]]
## Link
https://www.dropbox.com/s/7nidcmwdih6q1mr/the%20burden%20of%20knowledge%20and%20the%20death%20of%20the%20rennaisance%20man.pdf?dl=0

[Web URL for this note](http://notes.benjaminreinhardt.com/The_Burden_of_Knowledge_and_the_‘Death_of_the_Renaissance_Man’_Is_Innovation_Getting_Harder)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/The_Burden_of_Knowledge_and_the_‘Death_of_the_Renaissance_Man’_Is_Innovation_Getting_Harder)






# The tech tree is a scaffolding inside the knowledge frontier
If the [[Knowledge frontier]] is the surface of a high dimensional blob, then you can imagine the tech tree as the scaffolding holding the whole thing from the inside.  Any time that you add a new leaf to the [[§Tech Tree Model of Heuretics]] you’re expanding the edge of the knowledge frontier. 

This realization came from the question of “what does one technology superseding another actually entail?” It means that the technology has a larger impact. If the environment is fixed, larger impact means that it is higher than another on the relevant dimensions for that environment. ([[A heuretic’s impact is not endogenous to the technology but depends on the environment]])  

*That* in turns that we need to be looking at innovations in a massively multidimensional way. The *same* way that we look at the knowledge frontier. 

This model gives a new way of looking at [[§Type I and Type II Progress]] because Type I progress starts to look like progress endogenous to the tech tree, while Type II progress looks like expanding into a new environment. 🤔



[Web URL for this note](http://notes.benjaminreinhardt.com/The_tech_tree_is_a_scaffolding_inside_the_knowledge_frontier)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/The_tech_tree_is_a_scaffolding_inside_the_knowledge_frontier)




Jason Criticisms on Type I and Type II progress

# Type I and Type II Progress - publishing version
### What are Type I and Type II Progress and why should you care?
Imagine human capability as a high dimensional blob.[^1]  There are some people who live on the surface of the blob (let’s call it the [[Knowledge frontier]] ) - they have access to the best of the knowledge and resources humanity has to offer. *Pushing this frontier and expanding maximum possible capabilities is Type I Progress*.
There are many people who don’t live on the surface of the capability blob - most people who live in ‘developing’ nations, for example.  *Actions that expand these people’s capabilities towards the frontier are Type II Progress*
As William Gibson said [“The future is here — it’s just not very evenly distributed”]([William Gibson - Wikiquote](https://en.wikiquote.org/wiki/William_Gibson)) You could think of Type I progress as expanding what the future is, and Type II progress as distributing that future. 
There are even some people who might live a little beyond the surface in some dimensions. These are the innovators, the creators of new knowledge. 
People often confuse rapid Type II progress for Type I progress. This is in part because [[Progress frontiers are geographically fractal]] (more on this later) and as a consequence [[Type I and Type II progress don’t correspond 1:1 with developed and developing nations]] . However, [[You can only push the frontier on dimensions where you are at the frontier]]  and as a consequence, [[Frontier-dwellers tend to live in developed places]]
It’s important to note that [[Wealth redistribution is not type II progress]]
To frame the difference between the two sorts of progress: In [[Lever of Riches]] Mokyr writes: `“At any moment there is a large gap between average and best practice technology; reducing this gap by disseminating the techniques used by producers at the cutting edge of knowledge is technological progress without invention.” `  This process of gap reduction is Type II progress. 
And to put it graphically, here’s a plot of how agricultural output per hectare has changed over time. 
![](BearImages/7DC1F1A3-486C-4CA3-888A-22296E8235E4-792-00020D67BFA98E6B/9D76763E-DA30-43B3-B336-8344A32A5A94.png)
Why does this distinction matter?  Many discussions about “progress” conflate the two or devolve to debates about which type of progress is more important: “How can we be building rockets when there are starving children in Africa?” But [[“Should we focus on Type I or Type II progress?” is the wrong question]]. [[Debating the abstract relative merits of Type I and Type II progress is a waste of time]], and I suspect that disentangling them will lead to much more productive investigations about progress.
### Progress Frontiers are Geographically Fractal
There are parts of China (Shanghai, Beijing) that are expanding human knowledge and capability - building new augmented reality technology, CRISPRing babies, etc. But then there are also parts of China that don’t have indoor plumbing.  So clearly  [[Type I and Type II progress don’t correspond 1:1 with developed and developing nations]].
In the United States you would expect California to be right on the knowledge frontier - it’s the wealthiest state with the most progressive policies. Yet there are still tons of farm workers who live lives comparable to residents of developing nations - they clearly are not at the frontier and would benefit more from Type II progress than Type I progress.
Zoom in farther to the San Francisco Bay Area. In only a few square miles there are people pushing the limits of quantum technology and infrastructure and policies that look like something out of a developing nation.[^2]
You can’t even say that individuals are the quanta of where type I vs. type II progress happens. Many researchers on the cutting edge of science follow dietary guidelines from the 1970’s. 
Fractals are multi-scale patterns. Every time you zoom in on a fractal, it has the same pattern but on a different scale. This inability to cleanly separate the two types of progress geographically adds evidence to the fact that  [[“Should we focus on Type I or Type II progress?” is the wrong question]].
This fractal nature should make it pretty clear that
### Type I and Type II progress don’t correspond 1:1 with developed and developing nations
But if you’re not convinced, let’s look at the counterfactual statement: “Type I and Type II progress correspond 1:1 to developed and developing nations.” 
There are parts of China (Shanghai, Beijing) that are expanding human knowledge and capability - building new augmented reality technology, CRISPRing Babies, etc. But then there are also parts of China that don’t have indoor plumbing. 
There are even parts of the United States that do not have indoor plumbing or ride-sharing, etc.
However, the inability to say “developed nation, therefore Type I progress” and vice versa isn’t to say they’re uncorrelated. *Type I progress is easier to do in developed nations.*  It is just harder to spend time advancing the future when there are fewer people with all the knowledge and abilities that are at the frontier.  Additionally, new tools, processes, and devices often require previous infrastructure which is rarer in developing nations. This suggests that developed nations should prioritize Type II progress.
To put it differently
 
### You can only push the frontier on dimensions where you are at the frontier
Many people cite countries in Africa and China “technologically leapfrogging” the United States.  Two big examples are Chinese mobile payments and cell phone propagation in Africa.  These places can push the frontier *on those dimensions* - mobile phones as the computing center and mobile payments. However, this is not to be taken (as many people do) as a sign that they will soon blast past in other dimensions regions that lag on those along those dimensions unless there is a *precise* reason that for example mobile payments will enable better material science.
If you accept that you can only push the frontier on dimensions where you are on the frontier, it leads to another possible line of inquiry: asking the question, ‘which place or group is best set up to make progress in this specific dimension?’ 
### “Should we focus on Type I or Type II progress?” is the wrong question
Because [[Progress frontiers are geographically fractal]] you need to specify the scale and location you’re talking about. Once you have done that you can ask “at this scale and location what would be the result of a focus on Type I vs Type II progress?”  For example [[Type I progress is easier to do in developed nations]].
Organizations and institutions can then decide where they should focus based on their specific scale, location, and mandate. For example this line of reason would suggest that a sub-saharan African nation should not be spending money on a particle accelerator.
Individuals should decide whether to focus on the type of progress that fascinates them the most because [[Debating the abstract relative merits of Type I and Type II progress is a waste of time]]. 
### Type I and Type II progress are separate but equal
“Ok, but really if you had $1B and you could only spend it on one, which should you do?” 
We actually don’t know, and the fractal nature of Type I and Type II progress suggests that there might not be a “right” answer.
You can make arguments both ways: 
Over a long enough timescale if you focus on Type II progress, you will have more Frontier-dwellers.  Because[[Frontier-dwellers tend to live in developed places]] and [[You can only push the frontier on dimensions where you are at the frontier]], more people pushing the frontier will lead to more frontier expansion. 
However, the rate of Type II progress may be proportional to the distance to the frontier, and at the very least  [[In the long run Type II progress cannot exist without Type I progress]].
### So what?
If you buy these arguments, here are the actionable takeaways:
- Start distinguishing whether you’re talking about Type I or Type II progress.
- Be precise about the scale, dimensions, and context over which you’re talking about progress.
- Decide which one you’re more excited about and work on it because they’re separate but equal.
[^1]: According to a mathematician friend, the correct term for this is a “compact” but for  the sake of evocative language I’m going to stick with “blob.”
[^2]: Some time-dependent evidence of this are the PG&E blackouts in October 2019. 

See [[Type I and Type II Progress Publishing V3]]
---
References:
[The Illustrated Guide to a Phd](https://www.notion.so/innovate/The-illustrated-guide-to-a-Ph-D-e4b0ed9ae1b74de9b8ee981f27db381d)
[Productivity Growth in Agriculture](https://www.notion.so/innovate/Productivity-Growth-in-Agriculture-An-International-Perspective-Google-Books-6f2195eda23d40f8ba21890fcc189201)
[[Lever of Riches]]
             See [[Type I and Type II Progress Publishing V3]]
---
References:
[The Illustrated Guide to a Phd](https://www.notion.so/innovate/The-illustrated-guide-to-a-Ph-D-e4b0ed9ae1b74de9b8ee981f27db381d)
[Productivity Growth in Agriculture](https://www.notion.so/innovate/Productivity-Growth-in-Agriculture-An-International-Perspective-Google-Books-6f2195eda23d40f8ba21890fcc189201)
[[Lever of Riches]]
             

[Web URL for this note](http://notes.benjaminreinhardt.com/Type_I_and_Type_II_Progress_-_publishing_version)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Type_I_and_Type_II_Progress_-_publishing_version)




# Understanding the US National Innovation System - Paper
`In the conventional view, innovation is something that just takes place idiosyncratically in “Silicon Valley garages” and research and development (R&D) laboratories. But in fact, innovation in any nation is best understood as being embedded in a national innovation system (NIS). `

“Success Triangle” 
- business+society on one side, 
- trade+tax+regulation,
- innovation policy environment
::2/3 of these sides are government …::

1890’s first time US got to the [[Knowledge frontier]]
`As business historian Alfred Chandler showed, the large American market enabled U.S. firms to successfully enter new mass production industries, such as chemicals, steel, and meat processing, and later autos, aviation, and electronics.4 Because scale mattered so much to innovation and firm competitiveness, U.S. firms like DuPont, Ford, GE, GM, Kodak, Swift, Standard Oil, and others became global leaders. `

US didn’t have to overcome craft-based industries as much as europe

Government
- Funded infrastructure
- Land grant colleges→ agricultural research→ agricultural productivity
- Antitrust acts in 1890 and 1914

Mentioned Charles Morris’s *The Dawn of Innovation: The First American Industrial Revolution*s 

Large corporate labs started in 1930’s
US R+D labs started after WWII ::What was the decision makig here::

The CITP program (1963) was to provide funding to universities to do research helping innovation in sectors thought to help society, such as coal production, housing, and textiles. But didn't make it through congress.

`The new program, the State Technical Services program, was to fund university-based technology extension centers in the states that would work with small and mid-sized companies to help them better utilize new technologies ` ::Any time you want to get $$ through the federal syste, you need to appease the states::

`These efforts were followed up by efforts by Congress and the Reagan and Bush I administrations. Indeed, policymakers responded with a host of major policy innovations, including passage of the Stevenson-Wydler Act, the Bayh Dole Act`::[[The relationship of government, academia, and technology fundamentally changed in 1980 with the Bayh-Dole act]]:: `, the National Technology Transfer Act, and the Omnibus Trade and Competitiveness Act. They created a long list of alphabet soup programs to boost innovation, including SBIR (Small Business Innovation Research), NTIS (National Technical Information Service—expanded), SBIC (Small Business Investment Company—reformed), MEP (Manufacturing Extension Partnership), and CRADAs (cooperative research and development agreements). They put in place the R&D tax credit and lowered capital gains and corporate tax rates. They created a host of new collaborative research ventures, including SEMATECH, the National Science Foundation (NSF) Science and Technology Centers and Engineering Research Centers, and the National Institute of Standards and Technology (NIST) Advanced Technology Program. And they put in place the Baldridge Quality Award and the National Technology Medal. `

SV things made the government less worried about innovation but now worried about manufacturing

(7)
- Business Elements
	- Managerial talent is good
	- Short term time horizon pressures
	- Computer adoption
	- Venture Capital is strong
	- Wide array of financing available, limited government financing
- Cultural Factors
	- Demanding consumers
	- Culture of “Yankee ingenuity” - tinkering, inventing, making things better. Questioning things. Failure not dooming career.
	- 19th+20th century belief in progress but that has changed in 21st c. (9) 
	- Innovation clusters
	- In the past, high rate of investing in R*D
		- Govt 1960’s - 2.8% of income into R+D. 2010’s - 0.48% into R+D
- Regulation
	- Relatively lax antitrust
	- Relatively easy to start business but not the easiest
	- Annoying levels of regulation
	- Pretty high transparency and rule of law
- Tax Trade Economic Policy
	- Tax policy - favors R+D, high corporate taxes
	- Trade Policy - relatively lax
	- IP - constitution gave congress power to promote science and useful arts. Dominant opinion that without IP people would create less. Possibly too liberal in issueing patents
	- Standards - ANSI: voluntary standards for private firms
- Research and Technology
	- Mission Oriented research @ federal labs, curiosity driven research at universities
		- $140B R+D in 2013
		- Government R+D  decreased as % GDP
	- Federal labs - between 80 to 100
		- For the most 
	- University research 
		- NSF funds research unrelated to agency mission goals
		- Based on linear model of research
	- Tech Transfer System
		- Federal policy ignored tech transfer before 1980's
		- Stevenson-Wydler Technology Innovation act - 1980
			- changes to better enable the transfer of technology from federal laboratories to commercial use 
		- Bayh-Dole act enabled universities to retain IP rights
		- Tech Transfer Act 1986 etc. ::Dig into this:: (16)
		- NSF and NIH began pilot programs to link funded research to commercial outcomes
	- Research in Business
		- Tax credit
		- SBIR
		- anti trust exemptions for R+D consortia
		- National Network for Manufacturing Innovation - add
			- Four centers?
	- Systems of Knowledge Flows
		- Innovation clusters - state, not federal levels
		- US has relatively developed University - Industry collaboration system
			- `For every MIT or Stanford, there are 10 universities where commercialization is more haphazard and less effective. `
	- Technology Import/Export
		- Tech import via foreign direct investment
		- CFIUS monitors acquisition of US companies by foreing entities
	- Tech diffusion and Adoption
		- Department of Agriculture helps farmers adopt best practices
		- Smaller efforts to help manufacturers do similar - MEP
	- K-12
		- Pretty meh
		- Diverse schools
	- Higher education
		- More and more foreign PhDs 
	- Skill training
		- Private or on State level
	- Immigration
		- Immigrants 15-26% of new companies in US, 40% of new technology cos
		- Recongnition that system needs serious reform

Problems
- Lower funding for R+D
- Shorter term equity investing
- Neo Ludditism
- Bipartisan support for high skilled immigration but opposition from left/right extremes

### Published
2014
## Questions
## Thoughts
## Related
25. Matthew Stepp, Sean Pool, Jack Spencer, and Nicolas Loris, “Turning the Page: Reimagining the Nationals Labs in the 21st Century Innovation Economy” (ITIF, June 2013), 

## Authors
[[Robert Atkinson]]
## Link
[Understanding the U.S. National Innovation System by Robert D. Atkinson :: SSRN](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3079822)



[Web URL for this note](http://notes.benjaminreinhardt.com/Understanding_the_US_National_Innovation_System_-_Paper)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Understanding_the_US_National_Innovation_System_-_Paper)




In the same way that species don’t suddenly replace each other, technologies and ideas don’t suddenly replace each other

A theory that is not fully correct can still be effective

# Uptimal solutions in science and technology are rarer than you think
Take Newtonian mechanics and Einsteinian mechanics: Einsteinian mechanics describes the universe strictly better than Newtonian mechanics. Einstein can predict anything Newton can predict but Einstein can predict things Newton cannot. Einstein is “less wrong” than Newton in this sense. However, people still use Newtonian physics *all the time* even though they know it is ‘wrong.’ Why? Because it still gets most things right and is way easier to use.

We often look at only one parameter of a science or technology and as soon as a new idea supersedes on that parameter we declare the old thing obsolete. This single-parameter obsolescence is especially prevalent in science. The ideal in science is that as soon as a theory has strictly more powerful explanatory power than another theory it should completely replace that theory. (See [[Karl Popper]]) However, the reality is that we evaluate theories on more than just the parameter of explanatory power. Multi-parameter evaluation is one reason why myths persist - they are very transmissible or rememberable.  

[[Outdated science theories and outdated technologies can both remain on the Knowledge frontier]]!

The ‘non-primary’ parameters of a heuretic affect its dispersion. [[Dispersion is the process of a heuretic impacting the world]]. If a theory is both extremely predictive and hard to transmit, not only will it disperse more slowly but it’s “carrying capacity” in the environment will be lower. ([[Constraints limit the impact and dispersion speed of a heuretic]].) 

### Related 
- Inspired by [[neumannOneProcess2020]]
- [[Knowledge frontier]]
- [[§Tech Tree Model of Heuretics]]
- [[Technodiversity is important]]



[Web URL for this note](http://notes.benjaminreinhardt.com/Uptimal_solutions_in_science_and_technology_are_rarer_than_you_think)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Uptimal_solutions_in_science_and_technology_are_rarer_than_you_think)




Physical experiments are expensive

Arguments for focusing on Type II progress

Most people don’t drive Type I progress

# You can only push the frontier on dimensions where you are at the frontier
Many people cite countries in Africa and China “technologically leapfrogging” the United States.  The flagship evidence is Chinese mobile payments and cell phone propagation in Africa.  These places can push the frontier *on those dimensions* - mobile phones as the computing center and mobile payments. 
However, this is not to be taken (as many people do) as a sign that they will soon eclipse the US technology. [Examples needed.] Precision is essential when we’re talking about the technological frontier.
At the end of the day [[Type I progress is easier to do in developed nations]].



[Web URL for this note](http://notes.benjaminreinhardt.com/You_can_only_push_the_frontier_on_dimensions_where_you_are_at_the_frontier)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/You_can_only_push_the_frontier_on_dimensions_where_you_are_at_the_frontier)




Benʼs working notes

Knowledge frontier

Links to this note