I'm really curious how much effort there was in building up the data set - before, training the model before you got to "music"
Reading the steps feels like 9 months to a year before you got to credible music.
What kept you going in the belief this would work. I can think of 20 reasons why this shouldn't work - hence its "surprising" that it does. Its quite easily something you could have worked on for 5 years with no results.
reading your background - it also sounds like your time would be tightly constrained hence figuring out where to deploy it - you need to have some conviction you'll have success
- Loads of account details available - US, AUS, USD, EUR
- Super cheap all of the time (think revolut is cheaper for some amounts and routes)
- Plastic card
Hey Andy — love this. But really you are developing a framework for figuring out how much time to invest in research.
A couple of questions — which was prompted by your blog
What situations is there no value in research ?
Corrollary — are there ways you can systemically reduce your risk by increasing your gut feel for probabilities without reesearch ?
Final one — you make the great point on unknown unknowns — apart from systemically working through all possible states and thinking through various envrionmental changes — are there any systemic approaches to reducing this risk.
In working with startups — I would summarise an alternative approach that covers all 3 by the following pieces of accepted wisdom from startups.
Launch early — to validate in the market with real people
Talk to customers to build an empathy or gut feel for relative priorities of stuff — and what really matters
I think these two principles go some way to solving the problems you are highlighting through a different approach.
I’m sharing — as I’m sure you have considered this appraoch — and have a reason for discarding it.
Your post goes to the point at the heart of philsophical number theory.
What does equality mean ?
Yup – you got functinal equivalence, isomorphism, and temporary assignment of values.
But I think you could prove – that all these types of equality – are “instances” of “different implementations” of “equivalence.
They are no more equivalent than 1 = 1 is equivalent.
I.e. 1 = 1 means I think we can define a bijective “counting function” that proves there’s the “same number” of “elemetns” in the “sets”
I think (not sure) – if you define – counting fucntion / same number / elements / sets differently – you get the differing definitions of equivalence you enumerate.
The interesting thing for me is that 1 = 1 is defined clear in 4 of peano’s axioms
And you could mentally – try to develop different (and potentially) – more powerful notions of “equivalence” – with differing axioms
A final point… the prevalence of several “similar” concepts of equivalence in computer science – may point to an underlying “platonic idea” of equivalence – that either exists dormant in the world awaiting for us to discover it; or is a useful “technologocial” construct – that has accelerated “progress”
1. The problem with financial services is customers are greedy - not banks
2. The challenge in building any business - is that they always start by trying to solve a problem - but power always corrupts
1. If google’s aim was to truly organise the worlds information or xxx would it behave like thsi
2. We say capitalism is good as through the pursuit of capital appreciation - these kinds of problems aggressively get solved but we reward financially - the creation of monopolies
3. This then leads to all kinds of things: styling of innovation; but also pronounces the accumulation of capital
4. The solution is whether you can construct a company in another way
5. I think its now clear that you strive to be too greedy - that power / money will eventually corrupt
1. Google search results
2. Uber vs lyft
3. Microsoft
6. But this process takes years
Building financial services in a way that doesn't corrupt is as hard / harder than building a serach engine / browser / car sharing company that will last for 1000s of years rather than decards
The problem with financial services is customers are greedy - not banks
2. The challenge in building any business - is that they always start by trying to solve a problem - but power always corrupts
1. If google’s aim was to truly organise the worlds information or xxx would it behave like thsi
2. We say capitalism is good as through the pursuit of capital appreciation - these kinds of problems aggressively get solved but we reward financially - the creation of monopolies
3. This then leads to all kinds of things: styling of innovation; but also pronounces the accumulation of capital
4. The solution is whether you can construct a company in another way
5. I think its now clear that you strive to be too greedy - that power / money will eventually corrupt
1. Google search results
2. Uber vs lyft
3. Microsoft
6. But this process takes years
Ha - @TransferWise... we approach this slightly differently.
Most of the discussion here is about - setting to price to optimise for an objective. e.g. how to optimise for revenue, profit - or cash that you can invest in marketing.
Ultimately all these approaches seek to optimise for growth - depending on the definition / time frame / risk profile.
So what you think about growth as an OUTCOME, not an OUTPUT of the model.
We describe ourselves as a mission driven startup - with the mission of making the worlds money move at the touch of the button, instantly for almost nothing.
We invest in making international payments faster, cheaper and less of a pain.
As we reduce our costs, our price drops (obviously money isn't free and we need to have a small margin to cover our costs and continue to invest in the platform)
Hence - there isn't anyone here thinking, lets drop price by 1% and see if we get 1% more volume. We've built a conviction that if we continue to invest in aggressively dropping price customers will switch to us.
Note - if we approach this in a very data centric way will not move with the speed and aggression we are on this. Also our authenticity on this mission would be questioned by our customers.
This authenticity, and conviction - on not focussing on maximising the amount of value we can extract from our customers - is what driven our Word of Mouth growth rate. More on this here - https://www.slideshare.net/pnilan/slides-from-jam-london
So the reality is that - "moving money" costs - almost nothing.
Think about it. The process of doing a bank to bank transfer (what transferwise does) - involves one bank in one country, crediting another banks account in the originating country, and then the money turns up in a "correspondent"or "nostro" account in another country. The matching bit helps - but the real magic - is that monopolisitic incumbents have been charging spreads for years - TransferWise is different as they aren't greedy.
You are right - that fraud and KYC checks - are significant in the industry - but lets step through what these are.
I can share that we actually have pretty sizeable complaince and fraud teams - larger than many established financial services companies - but unlike them Fraud and KYC is part of the product - not a sign off team. The team understand our global regulatory responsibilities, and try to understand how tomeet them; whilst trying to figure out how to reduce customer and friction andcostin the process. There's been a complete lack of innovation in this domain for decades due to the dynamics you've outlined.
Forexample with KYC:
From KYC perspective - you have to verify that a customer is who they say they are - and check they aren't on any AML blacklists etc.
Different jurisidictions - have different thresholds (limits) at which they ask for KYC checks in place.
With Western Union Digital (and with transferwise) - someone may take apicture of their passport, driving licence, scan and upload. This then needs to be verified.
For its first few years - like other players in the industry - validated every single one of these documents individually - but over time got to a point where they had built up enough data to be able to validate these documents in an automated fashion. Inevtiably given how strict regulation is in this industry - the onus is still not on letting the bad guys through
I can't comment in public on transferwise's profitability - but can say the business has been operating sustainably for over a year - i.e. not losing money on a transaction - and investing sensibly on marketing.
Marketing is relatively small spend for Unicorn B2c startup - with over 80%+ of new customers coming in through WoM
The borderless bank account seems available in a ton of countires too - https://transferwise.com/gb/borderless/