my reaction to the QED manifesto

The QED manifesto states a moral vision for the field of mathematics.  The vision is for moving from human-based knowledge management towards machine assisted knowledge management. The process initially relied on oral storytelling, then use of paper, and now a computers, for sharing insights enables progress.

Claim: the vision of QED can be realized in Physics now.

• Now because technology has decreased the cost of implementation.
• Physics because experiments are the arbiter of validity. 

What remains the same is individual and institutional incentives.


Physicist have historically gotten away with human-mediated storytelling of quantitative relations. What we currently use (.tex) is a machine parsable version of oral transmission. Comprehension does not scale well given the complexity of the subject.

One outcome is that if physicist continue with the current approach of human-mediated storytelling of quantitative relations, we risk selecting smaller and smaller sets of individuals capable of advancing the field, as well as requiring narrower specialization. Another outcome is decreasing transparency for external consumers.

To make progress towards the vision described in the QED manifesto, break the objective into tangible milestones. Each technical change is also a cultural change that requires individual participants to evaluate the trade-off of value. Technology change also can manifest as change to an interface/workflow.  Change can be minor or revolutionary. The more revolutionary, the more value is required. Value can be a result of creativity (Tex, doi) or investment of resources (arxiv). 

Milestones enabling the QED vision:

1. Symbol registry using Godel numbers. Improves reader's ability to compare content. 
2. Expression registry (presentation) using Godel numbers. Improves reader's ability to compare content.
   These two registries form a Physics catalog. The catalog lacks value beyond lateral comparison. Both registries help the reader, not the author. The author has no incentive to invest extra work that does not benefit them. (This is the tragedy of the commons.)
3. Connective explanation (story). An appendix for derivation. Author shows their work to ease the readers burden. Improves reproducibility
4. Interference rule registry (math)
5. Formal expressions registry (math) per presentation
6. Validation of steps 

Both formal expressions and validation requires significant effort by author or machine. Specific to and limited by selection of CAS. Again, value is mostly low. Does serve as filter for crackpot content. 

All of these milestones are present in the Physics Derivation Graph. This perspective of milestones is useful for segmenting the PDG barrier into constituent barriers. Each milestone has to have clear incentives for users to justify investment.

Registries are the opposite of (local) content dictionaries in openMath.
Model: doi, orchid

Each registry requires perpetual hosting and public accessibility.
Models: arxiv,

Individual researchers would only participate as authors of content if they expect that their investment results in a durable product that is used by the rest of their community. Their individual work also needs to be tied to employment incentive structures or reduce existing suffering. Reduction of individual suffering can be by increasing effectiveness or increasing efficiency.