immediately following FOSSAsia – see below.
Establish customer empathy for the early adopter segment – the non-US founder of a tech startup raising funding or setting up an ESOP.
Plan out critical path toward business success:
- web app product development
- funding either public or private
- research project: expressive language required for efficient support for international jurisdictions
- opensource rollout for both the code and the legals
- marginally profitable revenue collection
- for scale, all human interactions to be referred out to legal or consultant partners
We will answer these questions:
- Who else is working in this domain?
- Why haven’t they succeeded yet?
- How long will this take?
- What kind of organization do we need to build?
- What resources do we need to acquire?
- What deadlines?
- What conferences, if any, shall we attend?
- What core competitive advantage does the Legalese project have?
For each project participant, we evaluate:
- What level of commitment and contribution can we expect over the next 12 months?
- What level of compensation will be required to deliver that commitment?
Five days of afternoon sessions.
Each afternoon we hold two sessions.
Each session is two to three hours. Presentation, then discussion, then consensus.
invite guest in, together with any other users, to talk about why they found it useful.
guest will come at 10am monday morning to our resort venue.
history with SPF
customers: martin from ncinga, wobe
minefields and milestones
counterparty offers feedback, you reject half of it
inline or up front?
functional evaluation – looking things up, six bookmarks deep
when you leave, you leave with you
different teams don’t share.
rough consensus and running code.
what to do about confidential information?
a new contributor’s guide.
start with a readme file. not “come and talk to meng.”
then you see a CONTRIBUTING file in the repo somewhere, connected to issues.
mariadb uses jira. worklogs.
let’s consider using https://gitter.im/ which is more open than slack.
could this be a good project for google summer of code 2017?
http://hackingbusinessmodel.info/
dogfood: get paid to do a pull request.
post-payment = convertible note.
pre-payment = executory contract. because it locks the issue to you and others can’t work on it.
so we want to disrupt the incumbents.
let’s look at the state of the art – what other people are thinking.
Is this even a good idea at all? http://feld.com/archives/2010/04/failing-fast-at-standardized-seed-deal-documents.html
different families of approaches:
A brief pictorial overview https://drive.google.com/open?id=0BzJyMUr3UozTNTZvT3BJMHZUUVk
A more detailed look at the landscape including whether founders has law and CS training https://docs.google.com/spreadsheets/d/1UJzUk_ejxUioU276Da8dSebEF7vJFAi1_Yw1_vxSRGE/edit#gid=0
why we’re in the right place at the right time.
the MySQL / MariaDB story
pretend you’re a founder raising funding. put everyone in the shoes of the early adopter.
first, use it as a user.
then, install a dev environment and get set up as a developer.
overview of the issues list in github
should we build a DSL for smart contracts, or a DSL for contracts?
is there any way to shorten the ingredient list?
plans for the next 6, 12, 24 months.
in terms of marketing, revenue, and funding.
practice with issue creation and task allocation.
what do we do about simultaneous allocation?
practice with communication tools
practice with holacracy circles
practice with interacting with the outside world. porous boundaries.
What does and doesn’t require permission from others?
What should and shouldn’t require notifications to others?
In which everybody gets on to The Same Page.
Legalese v1.0 is the current Google Docs app.
With the six critical issues resolved, that makes it usable by a determined end-user, it becomes version 1.2.
Legalese v2.0 has a much nicer skin and front-end with defaults that pre-fills and wraps the v1.0 spreadsheets.
v2.0 handles, for any Singapore startup,
- incorporation
- NDAs / confidentiality
- ESOP
- adding a co-founder
- employment agreements
- volunteer agreements
- seed fundraising round
v2.0 includes unit tests to verify each of these workflows.
We have a simpler front-end that encapsulates the spreadsheet tier.
We charge money for the premium version of the basic product.
IP and CLA sorted out.
Legalese v3.0 has a working, very simple, DSL to English compiler, with some ugly hardcoded crap in there.
We have expressed at least three seed investment agreements in the DSL.
The three seed investment agreements compile to working templates.
Those working templates are available for actual use by the v1.0 system.
The DSL compiler can handle
- inter-document references and definitions
- to english
- to something other than english
The DSL is usable by a law firm to build its own proprietary templates. It can build the agreements, or we might support a third party consultant who wants to help the law firm build the agreements, but we would prefer not to be supporting the law firm directly on a consulting/services model.
We could define the codebase is noncommercial; if you want to use this for white-label purposes, or get support, we will negotiate a licensing fee. Inspirations: apple app store; hosted wordpress with a theme marketplace. Charge each time the thing is used in production, e.g. Neota Logic.
Legalese could then make money on hosting the front-end and charging third party vendors to be on the marketplace.
we shall have sold at least one workflow to a startup which may be JFDI or non-JFDI.
Prove the direct sale model to see if martin is willing to pay for the current work.
Charge for conversion of agreements to XML. Train staff to do that.
develop workflows and XML templates for AU startups. Launch to some AU startups. See if AU startups will pay. Possible docs include http://www.startmate.com.au/financing-docs
To get from v3 to v4, the IP-heavy product development R&D will be built with the assistance of academia.
We will output to English.
need an R&D team on this.
In the course of a single deal, during negotiations, deltas will appear from the counterparty; maintain a parallel branch that applies those deltas, at the text/string level, to that deal’s paper, even while the master branch templates continue to update.
But it would be even cooler if lawyers could submit patches in the high-level DSL.
After 2017, we will decide if we want to have:
v5.0: round-trip isomorphism from reading English and turning it into the DSL. If the government offers us $50M in funding to do a LegalTech Research institute and they want Computational Linguistics to be a component, and they want to know how we will spend the money, we will say that some of it will be spent on this.
Data61 has already done a bit of that.
Hack the above using simulated annealing in the opposite direction.
Who’s going to do what? What approximate range of commitments might each person be able to indicate?
Each of the following is a circle containing multiple roles filled by one or more people.
Each circle gets its own Slack channel.
- v1 XML template Import and Maintenance
- Meng and Jobchong as primary support. Alexis to review all the existing templates to see if they are fit for purpose in SG. Add docs as needed to support the M@C Test Case.
- v1 backend support
- Meng and Anuj
- v2 Web Front End UI Circle
- Sebastiaan plus designer. We need a pixel pusher. Has in mind a girl he found in a school somewhere. The last girl went off to the north pole. Maybe this one will flee to Antarctica. Please try to find one who will stay in the tropics. User experience will be advised by Alexis. Anuj also involved in devops?
- v3 DSL R&D
- Michal, Long, Meng, Yochi, Chiah Li, Anuj will act in a variety of roles: designer, compiler author, test user. This includes DSL to English ontology and natural language blobs. Conversion of The JB Test Case into formalism.
- v3E DSL R&D
- Compilation to Ethereum. Anuj & Virgil (on tech), Yochi (on market research and customer demand).
- Evangelist and External Interface
- Virgil and Meng.
- Interfacing with Academia in Singapore
- Virgil and Meng
- Interfacing with Academia in Australia
- Yochi to explore and send out feelers
- Inbound Marketing and Blogging
- Meng would like any content we produce to be original && true && useful. Yochi to write a white paper for a law journal about how DSLs may change the world.
- Business Team to position, brand, and strategerize the product
- customer discovery to lead all the rest. Wardley mapping, the competitor analysis, keeping track of other projects.
- Compensation Committee
- to design the pre- and post-funding credit algos. negotiate each person’s compensation. #finance. Who will be on this?
- suggested: Chiah Li
- accepted: Virgil
- accepted: Meng
- sebastiaan suggests asking the Buffer Girl to come on and help architect it.
- Investment Team to herd the cats toward a close
- Lev.B.Man to levitate in and be the Closer while everybody else waits outside with empty coffee cups. Virgil, meng, Alexis knows rich lawyers, Chiahli knows rich people.
Each circle above should have a lead link and should have its own slack chat.
We plan to raise S$2M by the end of 2016.
$2M will help us run for 18 months.
S$5.48–8M pre.
$1M of angel money, 250k * 4/5, from SG, AU, US
$1M of institutional seed / VC money, maybe US, UK.
$1M of matching government money on CRP/IAF or SPRING TECS POV – on reimbursement basis, so not really real.
ESOP will probably be 15%
But let’s try a more enlightened, 21st century, holacratic style approach to compensation involving bounty ideas, badge-based compensation, stack ranking, Valve, Buffer, Stripe, transparency.
We can also let people balance their cash vs equity split each month.
Everybody should be sufficiently compensated & appreciated to continue working on the project.
We can offer convertible debt plus a convertible sofa in Venice.
People who can work on those terms – and they don’t have to work full-time – can be compensated in the same ways as everybody else.
We might be able to offer a very small cash drip.
Prior to the big funding round we can do a small funding round from other members of the company.
There will be different mechanisms to issue credit for work done. We could issue “Monopoly money” credit, which then converts to cash/notes/equity. Some algo will be involved here.
Each contributor can choose their cash/equity split each month. There will be some algo around this.
Let’s all read the book(s) on that stuff, or get our SOs to read the book and explain to us what it all meant.
Also Valve handbook.
Alexis to canonicalize the books and communicate the full set and have her be in charge of making sure everybody is indoctrinated.
Everybody needs to be on
- Google Drive
- Slack
- [email protected]
- Github
Who will be in charge of onboarding and reboarding everyone? Alexis.
Everyone is empowered to be the public face of the company, but in delicate situation are advised to caveat that not everybody else in the company might agree and they can’t commit the company/project to do stuff.
Also, inform the rest of the company that you like talked to whomever, so that we don’t appear totally frazzlebrained. This is called “CRM”.
So before you interact with anyone, just search for their name across:
- Slack
- Google Drive
- Github
And if we have had any interaction with them in the past the search should return a result.
We are not organized enough to enforce CRM discipline.
UI designers.
Don’t commit anyone else to work that they haven’t agreed to.
Volunteers are welcome, but we need a mechanism to gracefully mute non-contributing volunteers. Badge-based compensation will help with this.
Make something useful, or independent activity
Perform follow-ups, emails, start executing action items from the week
invite special guest that we can summarize to; get opinion, hear what they think.
friendly VC to explain what metrics they would want to see from us before investing.
to make our language more useful around version 3 or 4, we have to tag our functions as ones which are performable in the smart contracts world vs those which are performable in meatspace.
can’t we discriminate actors which are online vs offline; the ones which are online are first-class agents.
one perspective: courts would think about this in terms of damages. if a court can order specific performance, it is a meatspace action.
smart contract needs a way to know what has happened in meatspace. the event log (and related things like ipfs)
multisig supports jointly appointed independent verifier.
#lawdev has the details.
afternoon off, let’s go to sentosa. calamari at tanjung beach club.
then marina barrage to see skyline.
free. Some financial assistance may be available for interns/students/etc.
To register for the summit: submit a pull request that edits this document.
sorted by alpha, please
- alexis
- byte (Colin Charles) for first 2 days?
- jobchong
- mengwong
- oatsandsugar
- anuj
- advancingdragon
- virgil?
- Chiah Li
- Daniela
Meng will present Legalese in one of the conference sessions.
- use of opensource in legalese. gripes about academic prior art not being opensource.
- legalese itself being opensource, and comparisons to other infrastructure pieces like mysql/mariadb; the demographic of the opensource contributor as being both a user and a developer of the software, often as an adjunct to the day job
- legalese being a facilitator of “smart contracts” and contracts generally that are themselves opensource – blurring the line between creative commons and opensource content.
- some amusing remarks questioning who owns the copyright in a contract
- the situation with precedents is a lot like the state of software before opensource
the inspiring theme here, perhaps, is that humans are the only animal to use language as a tool. this interpretation of humanity elevates poets, lawyers, and programmers.
working backward from this agenda, what could the MVP for the purposes of fossasia look like? it could, very primitively, support the expression of a contract in our DSL, which converts to english and to ethereum. that expression of the contract should be opensourced, in a way that definitively addresses some of the unknowns highlighted in http://www.adamsdrafting.com/downloads/Copyright-NYLJ-8.23.06.pdf