Bell Labs

What are the different roles in an innovation org?

Bell Labs was only profitable in conjunction with AT+T

Bell Labs and AT&T had timescale alignment

Bell labs had a budget of 2.8 billion 2020 dollars

Why try to emulate DARPA and not Bell Labs

§Bell Labs Model

Week of 22 Nov 2020

Labs Document Outline

# Bell Labs was not a business
Bell Labs didn’t actually make most of its money directly off of its technology. Its primary source of revenue was just a % of the money from the local operating companies and AT&T itself. It gave technology to Western Electric to manufacture and they in turn gave it to AT&T and the operating companies who then were paid and gave the money back to Bell Labs. The trick is that at no point in this cycle is anybody putting a value on the technology directly. Because of its monopoly status, AT&T would be getting paid regardless of how good its service was. However, the only reason AT&T could maintain that monopoly was because they had a <semi-implicit> agreement with the government that they would continue to improve their services. [[Alignment requires existential threats]]. 

When each piece of the puzzle was separated and needed to interact via contracts and prices, Bell Labs clearly couldn’t stand on its own as a business. The exact explanation of this is tricky and requires some detailed analysis on data that may not even exist.  

![](BearImages/E3753743-2D42-4C5F-902F-81CE9479EEBB-88049-0006A6BE36DA27BB/IMG_0250.jpg)
According to Peter Drucker’s analysis of the AT&T breakup, it is possible that Bell Labs *could have* become a business, but that would have required radically reconfiguring from what it had been. 

Bell Labs didn’t actually produce products. Once technology was at a point where it was ready to be manufactured at scale, it was handed off to Western Electric. If both organizations weren’t wholly owned subsidiaries of AT&T,  Western Electric would both want a cut (reducing Bell Labs’ hypothetical profit margins) and would either need to be paid up-front or convinced to manufacture new things that they would then sell to AT&T.

It wasn’t that Bell Labs wasn’t creating value - it’s just that it was doing it in the context of [[“The System” of AT&T]].   
![](BearImages/6DE120B4-C48E-48D4-9713-D49775DDC3BA-88049-0006CEF5E8D607A5/IMG_0231.jpg)

Perhaps one way to look at it is through the lens of [[Transaction costs]]. Bell Labs could only exist because it was getting continuous payments for extremely lumpy outputs (presumably some years, the direct value it delivered for AT&T was massive and some years it was near zero. Because of how human brains work, few companies would be willing to sign a contract that traded a consistent cash flow now for an uncertain innovation in the future. There are also some things that cannot be written into contracts - the trifecta relationship between Bell Labs, AT&T, and government regulators was incredibly valuable to AT&T but couldn’t be made explicit and wouldn’t have been possible if Bell Labs were a separate entity. 

It’s important to note that Bell Labs was not a business both to build a mental model of the organization and to stare a major problem with trying to build the next Bell Labs straight in the face. [[Healthy corporate R+D requires a trifecta of conditions]] and it’s unlikely that those conditions will arise outside of software any time soon. People have implicitly realized this and most ideas to ‘build the next Bell Labs’ involve creating an independent organization. However, building an external organization will require money and  [[People don’t invest in technology, they invest in businesses]]. The upshot is that any spiritual successor to Bell Labs will need to have a radically different structure than the original. 

This note is perhaps a dual to [[Bell Labs was only profitable in conjunction with AT+T]]. 

### Related 
-
- [[The Nature of the Firm]]
- [[The Idea Factory]]
- [[Research should be treated like infrastructure]]



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

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




Top of mind

Weekly Planning Week of April 26

Week of 26 July 2020

Michael Filler

Training Plan

# Week of 16 Nov 2020
This week I would ideally like to have a hypothesis about who I would want to pull into both a tele robotics and a simulation for APM workshop. 

### APM
On the  requires narrowing the search to something more distinct than “deep learning can help do simulations of things that involve the wave function. I’ll need to pull on the threads around Julia, PDE solvers, and be much more specific around what I actually want to do. One track of this is having some ignorant conversations with people. 

Perhaps run APM by [[Michael Filler]]

Another track will involve at least skimming and understanding [[Nanosystems]].

I also need to incorporate Eric’s feedback into the [[APM Roadmap]] but while doing so, I should meta-observe what I’m doing and try to suss out the “update procedure.”

### Telerobotics
On the telerobotics side, it means at least understanding the landscape of people who are at the frontier, who is good, and what their constraints are. This is much less about understanding the problem technically than trying to feel out who is truly excited about tele robotics as a vision for the future. I want to understand why it’s not possible to have a really good tele robotics simulation - is there actually a hard technical problem or is it just the sort of thing that the people who are building simulators don’t have the incentive to do. 

Part of this is getting the [[Telerobotics Roadmap]] in both a textual form and its block-diagram dual. 

Perhaps some of the writing time should be devoted to doing something along the line of [[bushScienceEndlessFrontier1945]] or [[Man-Computer Symbiosis]]. (Is there a term for this sort of writing? Speculative technical writing)

### Roadmapping
Throughout the process of APM and Telerobotics I should be paying attention into how to incorporate new information into the roadmap as I talk to people.

Some random thoughts
- Use cards during conversation to remind you of things?
- Assign each element of the roadmap a “certainty number” and direct conversations towards those? Maybe what you want is a precision rating.

[Web URL for this note](http://notes.benjaminreinhardt.com/Week_of_16_Nov_2020)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Week_of_16_Nov_2020)






Week of 29 Nov 2020

# Week of 22 Nov 2020
### APM
Winning looks like 
- Actually being able to articulate the arguments about why molecular machines shouldn’t work
- List of *precise* questions or *why* they’re imprecise
	- Slash goals
- Reach out to all the Feynman prize people
	- This requires reading the papers

### Telerobotics 
- Actually get the [[Telerobotics Roadmap]] written

I think in both of these, I need to pay attention to where my confusion is. 

### “Completed”
- [[Bell Labs was not a business]]
- [[Why try to emulate DARPA and not Bell Labs]]

#### Previous week
A lot slower than I was 
- [[Week of 16 Nov 2020]]



[Web URL for this note](http://notes.benjaminreinhardt.com/Week_of_22_Nov_2020)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Week_of_22_Nov_2020)




Week of 7 Dec 2020

# Week of 29 Nov 2020
This is supposed to be a week focused on the meta but … There are two major things that need to be done that are basically an extension of the thought-work from last week. It *may* be possible to get these both done on Monday - if I can do that it will be a huge win.
- Massive amount of reaching out to people. Do what must be done. Show no mercy. Do not hesitate. Fear is the mind killer, fear is the little death that brings total annihilation. 
	- Feynman prize people
	- Simulation people
- Vision writing 
	- APM
	- Telerobotics

Main: *Meta meta* Need to start gearing up to publish a manifesto thing in the new year. This week I want to step back and come up with a strategy of the pieces that is going to include, where I feel strong (ie. Incentive analyses), where I feel weak (ie. Roadmapping strategies, but then that can be called out as a discipline that needs to be developed), which pieces are core and which pieces are peripheral, what I want to do in terms of formatting, and the timing of releasing things. 

- Weight at beginning of week
- Max pull-ups at beginning of week

### Days
- [[30 Nov 2020]]
### Finished
- [[30 Nov 2020]]
	- [[In the future people could use remotely-operated robots to effectively teleport to do physical work]]
	- [[A face-screen on an artificial body probably won’t be a substitute for either the convenience of video chat or the intangible connection of in-person interactions]]
	- [[Telerobotic technology could effectively the closest thing to stepping into a new body or ‘I know kung fu’]]
- [[1 Dec 2020]]
	- [[Telerobotic technology can effectively enable people to have superpowers]]
	- [[Can the ARPA model exist outside of the government?]]
- [[2 Dec 2020]]
	- [[Labs implicitly judge research through the sense of dollars per paper]]
	- [[Timesharing could be extended from digital infrastructure to physical infrastructure]]
- [[3 Dec 2020]]
	- [[Precise visions are more likely to happen]]
	- Today was actually good - while I didn’t finish many notes, I feel like shifting into “ok what do I need to do in order to publish in mid January?” Has really shifted my mindset and got me thinking about useful things.
	- Really need to stop MTGA during the daytime
	- More UFO reports
- [[4 Dec 2020]]
	- [[Inverting the normal relationship between an R+D organization and is money factory may enable new classes of activities]]
	- [[Bell Labs enabled free radicals]]
	- [[Reaper functions beat culture decks]]
	- [[Reaper functions dominate other incentives]]
- [[6 Dec 2020]]
	- [[There are DARPA-style ideas that DARPA doesn’t pursue]]
	- [[Good Simulations could be the ‘why now’ of roadmapping tools]]

### Pevious Week
- [[Week of 22 Nov 2020]] - fuck I feel like I got almost nothing done

[Web URL for this note](http://notes.benjaminreinhardt.com/Week_of_29_Nov_2020)

[Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Week_of_29_Nov_2020)






Ben's working notes

Week of 22 Nov 2020

APM

Telerobotics

“Completed”

Previous week

Links to this note