§PARPA

You can create a program design discipline that enables better research and development

The value of planning is inversely proportional to the cost of experimenting

Program design could be thought of as creating a causal graph of blockers and hypotheses about how to mitigate those blockers

Treating Science Fiction as case studies from a future history

Program design is like creating the lego instruction book

Program design is a coordination tool

Instead of planning not working as well as people in the past thought, as a society we’ve become worse at technological planning

How does money work in PARPA?

The distinction between areas, programs, and projects

Physical experiments are expensive

Workshops could be a way to start small building a private ARPA

What role does scenario planning have in technological program design?

Maps help you figure out possibilities

There was a high return to technological roadmapping compared to not roadmapping in the past

Enabling Mission Impact - Funding Strategies for High-Risk High-Reward Innovation

How do you measure real progress in research?

Top of mind

An ontology around peoples attitudes competency and roles is important for program design

An ARPA-like program could be the first step towards making general-purpose telerobotics inevitable

The tech tree model suggests that explicitly drawing out tech trees is important

Explaining things like I’m five helps avoid the Einstellung Effect

Funding is path dependent and funding curves matter

A predictably successful idea is composed of known unknowns that that a team can execute against

Planning to Start vs Planning to Finish

The crux question of program design is when to commit

What are the different kinds of bottlenecks?

Say ‘X is more important than Y’ instead of ‘X is important’

There are two kinds of program design conversations — expansive and focused

The role of a visionary is to bootstrap the new from the old

What is PARPA’s edge?

What does a modified PARPA Heilmeier catachism look like?

Roadmapping is related to the ability to create leading metrics for success

How do you think about the risk profile of PARPA programs?

Jim Matheson and Joshua Krieger conversation 14 Jul 2021

Work backwards with questions and work forward with answers

Joe Rollin conversation 10 May 2021

Precisely listing entrenched beliefs could be an important part of bottleneck analysis

Abstractly you want a good way of visualizing a bunch of chained vectors

Block Diagrams need to be in easily editable format

Geoff Anders conversation 7 Jun 2021

Geoff Anders conversation 9 Jun 2021

Getting people to talk about speculative research extensions is hard

Josh Greenberg Conversation 18 May 2021

Roadmaps should have statements like “if we gain the ability to do X then we should do Y”

# §Program Design
- [[What are the goals of the program design process?]]]::needs work::
- [[Program design is like creating the lego instruction book]]
- [[Program design could be thought of as creating a causal graph of blockers and hypotheses about how to mitigate those blockers]]
- [[The crux question of program design is when to commit]]::needs work::

### Hypotheses
- [[The first step of roadmapping is frontier mapping]]
- [[A roadmapping framework may make the ARPA model more effective]]
- [[Talking to experts can expose bad equilibria, the moves necessary to get out of them, and considerations for doing them like cost constraints]]
- [[You can create a framework for talking to people that makes it easier to generate a roadmap]]
- [[You could build tools to make program design more efficacious]] 

### What are roadmaps
- [[What are the properties of maps?]]
- [[Roadmaps have two pieces - the map and the path]]
-  [[Roadmaps are a theoretical construct to escape inadequate equilibria]]
- [[List of Roadmapping Examples]]
-  [[Most roadmaps suck]]
	- [[Most roadmaps are not execution oriented]]

### Why Roadmapping
- [[Instead of planning not working as well as people in the past thought, as a society we’ve become worse at technological planning]]
- [[Roadmapping could be thought of as backwards-chaining together a series of conception-only heuretics]]
- [[Roadmaps can enable faster progress than blind exploration]]
- [[Technological roadmapping can give an idea of where technological evolution can go and how to get there]]
- [[A roadmapping framework may make the ARPA model more effective]]
- [[Program design is like performance debugging a vision]]
And of course we need to look at the possibility that this is all BS
- [[The roadmapping process is heavily context dependent and there may not be a generalized roadmapping process at all]]

### Building better roadmaps
- [[Design for roadmapping system]]
-  [[Person targeted questions for roadmapping]]
- [[What are the eight steps of Wigmore analysis translated to science?]]
- [[Technical Planning Principles]]
- [[Good roadmaps work backwards from a goal]]
- [[Good Simulations could be the ‘why now’ of roadmapping tools]]

### Glossary
- [[The distinction between areas, programs, and projects]]
- Bottlenecks -> Roadmap -> Execution
- Language
	- Binding Constraints
	- Affordances
	- Dependencies 
	- Project
	- Program
	- Area
	- Causal graph of blockers
- Steps
	- Frontier Mapping
	- Finding the right people
	- Comparing different paths
	- Goal setting
	- Finding serious contexts of use
	- Different institutional ‘moves’
	- Respectability cascades
	- Spending curves

### References
- [[And Suddenly the Inventor Appeared]]
- [[Wigmore Guide - Paper]]
- [[The trade-off: a central concept for biomimetics]]



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

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




Uncertainty always involves risk but risk does not always involve uncertainty

Many knowledge generating activities are inherently illegbile

What does ‘working’ mean?

Institutions either make things more legible or act as trust brokers

Risk needs a metric

People have different and unknown caring money thresholds

Anyone on the knowledge frontier is a bit of a crackpot

Innovation orgs need a money factory

When should a program that smells like research be a startup?

Narratives are how we deal with uncertainty

People like funding winners

We need to disentangle technology from tech

Heilmeier — Licklider — Braben — Elon quadrants of Research Management

Physics risk vs. Engineering Risk

Productive Uncertainty

Podcast List

Knightian Uncertainty occurs when you hit the limits of theory

Glossary

Derisking is actually deuncertaintying

Kanjun Qiu conversation 18 Apr 2021

# Knightian Uncertainty
`”Uncertainty must be taken in a sense radically distinct from the familiar notion of Risk, from which it has never been properly separated…. The essential fact is that ‘risk’ means in some cases a quantity susceptible of measurement, while at other times it is something distinctly not of this character; and there are far-reaching and crucial differences in the bearings of the phenomena depending on which of the two is really present and operating…. It will appear that a measurable uncertainty, or ‘risk’ proper, as we shall use the term, is so far different from an unmeasurable one that it is not in effect an uncertainty at all.”`
- [[Risk, Uncertainty, and Profit]]

`“There is a fundamental distinction between the reward for taking a known risk and that for assuming a risk whose value itself is not known. A known risk is easily converted into an effective certainty, while true uncertainty, not susceptible to measurement.”` (Paraphrased by [Explained: Knightian uncertainty | MIT News | Massachusetts Institute of Technology](https://news.mit.edu/2010/explained-knightian-0602)) 

The inability to predict, even probabilistically, what will happen

Knightian uncertainty is anchored on the idea of imperfect knowledge about outcomes in situations where it is hard to measure those outcomes. 

A corollary is that [[Uncertainty always involves risk but risk does not always involve uncertainty]].

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

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






The two ways to enable action is through legibility and trust

Projects need to build narratives of perceived progress

What are the different flavors of uncertainty?

# Narratives are how we deal with uncertainty
People are constantly acting in the face of deep uncertainty. Arguably *most* of our actions look more like [[Knightian Uncertainty]] than situations with a finite number of outcomes that could, or do, have probabilities associated with them. If we needed to do an expected value or risk calculation for everything we would be utterly paralyzed. This does, in fact, happen to people — I’ve been overwhelmed to the point of inaction with the multiplicatively cascading possibilities for what I could do and the resulting potential outcomes. At this very moment I could be writing this, or I could be sending an email, or I could be sipping coffee, or lying on the floor with my eyes closed, or going for a walk, or calling a friend… 

One could imagine navigating this uncertainty like Twoface from Batman, flipping a coin at every decision point. You could also imagine finding the best approximation to a risk curve for everything you do and doing a back-of-the-envelope expected value calculation. Most of us don’t do either of these things, and yet we manage to act through the uncertainty pretty well. 

Instead, we act based on the stories we tell ourselves.  “This is the most important thing because X,” “I’m doing this because …” “I have no choice but to …” Heuristics are narratives — “X is true enough that it’s worth acting on.” *We act through uncertainty by telling stories that collapse the massive decision-space to either measurable risks or a raw narrative.* 

Accepting that most people are acting off of narratives most of the time might seem like a rejection of logic and rationality. You might be tempted to say “well we *ought* to strive to act more from evidence and logic.” or to throw your hands in the air and say “so the only thing we can do is to make shit up?” The former is the path to self deception — if you actually accept that real uncertainty exists, by definition you can’t create risk curves and therefore acting based on evidence and logic is fooling yourself. The latter is overly nihilistic. 

I’m belaboring this point because I suspect that if we acknowledge that narrative is the only mental tool we have to act through uncertainty, we might be able to improve it. *There may be more powerful ways of constructing narratives.* For example, in the uncertainty-fraught case of creating new technology, you might ask: [[What role does scenario planning have in technological program design?]]

### Related
- [[Philanthropies are forced to show consistent short term progress]] because they need to do constant narrative maintenance 
- [[Historical narrative is not scientific inquiry or engineering design but it is valuable]]
- Perhaps this is just an extension to the distinction between [[Systems Thinking and Story Thinking]]
- [[Science fiction is a powerful tool for transmitting science or economic ideas]]
- [[Good science fiction can enable more awesome sci-fi shit]]

### References
- [[A story is a lie and a story is true]]

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

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






ICT Outline

# Top of Mind Questions
- Are there better ways to  deal with [[Knightian Uncertainty]]?
	- [[Narratives are how we deal with uncertainty]]
- [[What does a modified PARPA Heilmeier catachism look like?]]
- [[When should a program that smells like research be a startup?]]
- - How doe we do better [[§Program Design]] and specifically fleshing out the idea that [[PARPA creates causal graphs of technology blockers, precise hypotheses about how to mitigate those blockers and then executes on those hypotheses]]

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

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






28 Aug 2021

24 Aug 2021

25 Aug 2021

26 Aug 2021

27 Aug 2021

About these notes

# Top of mind
(Updated 19 Jul 2021) 

The intention of this page is to serve the function of  [a /now page](http://nownownow.com/about)  - shamelessly stealing the idea of a notes-based now page from the indomitable Andy Matuschak.

Most of my time is spent executing on the plan to build a[new institution](https://benjaminreinhardt.com/parpa-2-pager.pdf) I outlined [in the piece I wrote on building a private ARPA-riff.](https://benjaminreinhardt.com/parpa). This entails:
- Finding [potential program managers](https://benjaminreinhardt.com/parpa#its_possible_to_find_people_who_are_both_willing_and_capable_to_be_program_managers) and working with them on the fist steps of [[§Program Design]].
- Simultaneously working on designing a few programs myself. Specifically [general purpose telerobotics](https://benjaminreinhardt.com/parpa#general_purpose_telerobotics) and first steps towards an [artificial ribosome.](https://benjaminreinhardt.com/parpa#positional_chemistry)
- Working out the nitty-gritties of the [legal structure](https://benjaminreinhardt.com/parpa#legal_structure_is_critical_to_institutional_capacity) and getting the entities incorporated. 
- Finding the money to fund these things. 
- Reputationally bootstrapping the whole thing: the actual work becomes easier the more seriously people take you, but in order for people to take you seriously you need to do good work. 

If you want to see PARPA succeed, here are [[Ways to Help PARPA]].

More abstractly,  a thesis I’ve been playing around with as PARPA’s ‘theme’: [[Materials and manufacturing underpin civilization]] and the corollaries  [[What processes can be drastically changed with new manufacturing tools and capabilities?]], [[New materials can radically change technological constraints and enable previously unthought of applications]]

The nature of trying to create the organization still raises a number of research-y questions: [[Top of Mind Questions]]

And of course, I can’t help occasionally dicking around with
- Creating a [[“This would be amazing BUT” list]] — would love more suggestions!
- Expanding on the idea of [[The linear innovation pipeline is more harmful than useful]]
- Thinking about [[Observational technological phenomena]] and the [[§Tech Tree Model of Heuretics]].
- Recording an occasional [Idea Machines Podcast](ideamachinespodcast.com)as a way to talk to fascinating people and share it with the world.

On the physical front, I feel like I’ve mostly got my incremental climbing/physical ability increasing [[Training Plan]] nailed down. The trick is recovery and not overeating.



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

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




Sarah Constantine conversation 8 Jul 2021

General-purpose telerobotics needs work that’s not suitable for a for-profit company

Jerry Neumann conversation 1 July 2021

Software startups can get away with doing research

How to think about the differences between FROs and private ARPA programs

Week of 30 Aug 2021

14 Jul 2021

7 Jul 2021

# When should a program that smells like research be a startup?
*Note about language. I’m going to be trying to use the word ‘uncertainty’ in the technical sense of* [[Knightian Uncertainty]] *that carries a lot more meaning than just “we don’t know.”*

We can say for certain some things about programs that smell like research. 
0. To some extent they are attempting to do something, usually discovery or invention-wise, that nobody has done before. 
1. They will not be able to generate cash flow directly from their main activities for some amount of time.  
2. As a result of #1, the activity will need some kind of money factory [[Innovation orgs need a money factory]].
3. That money factory (or multiple money factories) will need to put money into the program for some amount of time before the program produces a ‘working’ result — a good business, a technology that other people can use, an acquisition etc.  
4. That timescale has unpredictability about the timescale to get to a ‘working’ result. More uncertainty means more unpredictability on the timeline.
5. In large part because of the timescale unpredictability, there is unpredictability about the amount of money to get to a ‘working’ result. The cost is often tightly coupled to the amount of time it takes. More uncertainty means more unpredictability.
6. The time and money are necessary to address technology risk (a chance that the technology won’t hit a performance metric) and some amount of technology uncertainty (not knowing what the metric actually is). 
7. There is some amount of unpredictability about potential outcomes: what the form of the output will be, whether there is a market for that output, whether the output will be of a form that can capture value from that market, whether the program will organizationally be able to be a business that can do that value capture. 

If the program is housed in a new for-profit organization, ￼#1+#2+#4 means that except in rare cases, the organization needs to be a high-growth startup because in order to provide competitive time-discounted ROI on the money the company needs to do the work over a potentially long timescale, it needs to target massive returns. Being a [[Grahamian startup]] then subjects the organization to [[§Startup Constraints]]. 

One could imagine situations where a for-profit research program could avoid committing to be a high-growth startup. A for-profit could get away with targeting lower growth/eventual value if funding comes from non-dilutive grants or ‘friendly’ investment where the investors don’t expect to make as much return on their money as they could elsewhere. However, both of these options come with their own downsides.

Outliers dominate both the results and these conversations so I’m going to call out the elephants in the room:
- [[SpaceX]] (which is debatable — at first they weren’t even trying to do something nobody had done before). 
- [[Genentech]]
- [[Intel]]?
- A number of therapeutic companies 
The fact that the list is so short compared to the list of researchy organizations that have failed to deliver on their promises seems indicative. The ‘failure rate’ is even higher than the average for startups.  

At the end of the day, no set of criteria can definitively say “this project shouldn’t be carried forward by a new for-profit organization” except maybe “can the desired output be sold?” Though even then someone could point to profitable but unintended discoveries. The nature of [[Knightian Uncertainty]] means that it’s certainly possible to flip all heads in a row on a large-dimensional, non-stationary coin. However, I would argue that the more the unpredictability in each of the steps smells like uncertainty instead of risk ([[Uncertainty always involves risk but risk does not always involve uncertainty]]) the less it’s a good idea for the project to be carried forward by a new for-profit organization. 

The correct way to look at the question of whether a program is too researchy to be a startup is probably in terms of a decision tree.
- Can the desired output be sold?
- How much market uncertainty is there?
- Is there an objective ‘measure’ of what ‘working’ means? How smooth is it?
- Is the technology a modular product?
- What ROI would investors consider acceptable? (Reduce this number)
- How long will investors be willing to wait to get that return?
- How long are investors willing to go without seeing legible progress. 
- Does success require multi-point changes?

### Can the desired output be sold or unlock a product?
I you are setting out to create something that could conceivably become a product or enable one, the program probably should not be a startup. There are examples of accidental but profitable discoveries that happened in the course of pursuing something out of motivations besides profit — discovering a heart drug while studying tree frog poison in the amazon for example — but you might as well just buy stocks chosen by a random number generator. 

### Is there a smooth objective measure of what ‘working’ means?
One way to look at uncertainty is that there’s no definition of what ‘works’ means: this could be in terms of the technology itself (technology uncertainty) or in terms of how it becomes a good business (market uncertainty). Note that this is different than technology or market *risk* where you know what winning looks like but there are a lot of reasons why you wouldn’t get there. [[Uncertainty always involves risk but risk does not always involve uncertainty]]. Technology risk sounds like “there’s a chance that we can’t get the efficiency of this process above 35% but the back of the envelope says there’s nothing that should prevent that and we have a development plan to get there.” (A metric and a vague sense of a probability distribution over that metric). 

Creating new technology will almost always involve *some* uncertainty. Technology’s modular/combinatorial nature[^1] means that there will be uncertainty at some level of the hierarchy. Sometimes that uncertainty will propagate up to the top-level technology if it is an irreplaceable subsystem. However, my extremely hand wavy intuition is that [[You cannot have massive uncertainty in subsystems without it propagating upwards]]. 

### How smooth is the measure of what ‘working’ means?
The most smooth situation is one where many incremental changes can cumulatively move the needle on the metric. 

Talking about 

[[There is a tradeoff between an organizational culture that is good at addressing market uncertainty and one that is good at addressing technological uncertainty]].  As a result, market uncertainty amplifies the chance that a researchy startup will fail even beyond the normal multiplicative nature of multiple sources. 

Some questions that are proxies for the degree of technological uncertainty facing a project:
**How smooth is the technology’s fitness landscape? That is, how much could a single change shift outcomes?** 

**Do you know what it means to improve?**  

**Do you know how you will know that you know how to incrementally improve the system?**

### Related
- A specific subset here is  [[General-purpose telerobotics needs work that’s not suitable for a for-profit company]]
- [[What does ‘working’ mean?]]
- [[Uncertain things basically have infinite risks]]
- [[Knightian Uncertainty]]
- [[There might be many situations with intractable uncertainty or randomness in the highest order bits]]
- [[§Startup Constraints]]
- [[Jerry Neumann conversation 1 July 2021]]

### Braindump
One trick is that the non-stationarity of the coin flip means that if the people on the program know that they only get a set number of flips before they produce a working thing

Are we allowed to throw out “what about Tesla/SpaceX”
- There needs to be a lot of excess value creating from the research work 
- [[Startups need to conform to investor timescales]]
	- [[Venture capitalists need to fund things that increase valuation in a few years and return in 5-7 years]]
	- The average stock market return is 9.2% 
	- Average over 2010-2020 13.6%
		- $1->$6.77 over 15 years
		- In order to beat that by even 2x a startup would need to turn $100M of investment into $6.8B returned to investors, so valuation would probably need to be $10B
- Startups working on general purpose technology tend to get sidetracked
	- Magic Leap
	- 
- Manufacturing a physical thing that has never been manufactured before has not led to great results outside of electronics 
	-  [[Startups need to produce a scalable product]] 
		- “But what about Tesla and SpaceX” 
- Need an actual breakthrough, as opposed to just grinding on current paradigms 
- TRL of the overall system

### Subset
-

[^1]: See [[The Logic of Invention]]

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

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






Ben's working notes

Top of Mind Questions

Links to this note