Taxing Crypto and Jurisdiction Shopping; Flipside of Digital Wealth; Machine Vision Calamities (via Autonomous ?NEXT)

Hi fellow futurists -- here are our top 3 favorite thoughts. 

Taxing Crypto and Jurisdiction Shopping.  

The American 2017 tax-year is over, which means a whole bunch of people are panicking about how they should pay taxes on their cryptocurrency gains without becoming Wesley Snipes. To start off, we highly recommend some professional tax advice, as well as the Forbes and Bad Crypto podcasts on this issue. The short answer is that nothing has really changed since the 2014 IRS memo treating crypto as property, which means anytime you get in and out of a crypto asset, that is a taxable event. But let's broaden the conversation to jurisdiction shopping, which is the practice of companies choosing where to domicile entities (like funds) and do business to optimize their choice of law. Israel just proposed some favorable ICO tax rules, which would allow ICO proceeds to be legitimized. Many of today's crypto companies have an entity in Gibraltar, Singapore, Switzerland or Estonia. Why?

There are several reasons we've heard that cryptopreneurs go jurisdiction shopping: (1) taxable treatment of ICO funds raised, (2) ability to open a bank account despite a decentralized fund-raising, (3) capital gains treatment on crypto currency transactions, and (4) treatment of token distribution to founders and advisors. As tax professionals develop sophistication in the space, these become more and more important. 

ICOs today try to avoid equity tokens so that they do not run into securities law, with its registration rules and requirements. But that implies the ICOs are selling digital goods, subject to Income tax on the profits. To solve this, projects structure a foundation to receive the revenue as a grant. That didn't work so well for Tezos. Second, the bank account issue is why we see so many foundations start in Gibraltar, where AML/KYC standards are more permissive, and thus companies can open bank accounts with proceeds from contributions of thousands of anonymous people online. On capital gains, we will likely see massive confusion come tax season, in whatever jurisdiction. The most conservative reading is that any crypto-to-crypto exchange is a taxable event, and requires the payment of fiat on the capital gain. Similarly, a purchase with Bitcoin of a sandwich will trigger a capital gain (or loss) realization for the price of the sandwich -- so you'd be paying tax there too. Check out Cointracking as a possible solution to at least know what you owe.

The last piece is probably the most invisible to folks who have not started private companies before. These tax issues, especially around illiquid private stock, pop up all the time. Imagine a startup worth $100 million and you join as VP Product, and get 1% of the equity, which is not yet liquid. Well, you also immediately get a $450,000 tax bill, payable that year. So in early stage tech, there are many solutions for these problems -- like options, 83(b) elections for stock that vests, and so on. Crypto has none of this. If a founder wants to file an 83(b) election and pre-pay their taxes on worthless tokens, they can't because the tokens are not equity. Or similarly bad would be having to pay tax on allocated founder tokens valued at a Tezos valuation in cash upon receipt. So the traditional tools do not work, and the potential tax burden to individuals involved in the space is quite severe. That's one way to slow down the pace of innovation.

Source: Cointracking

The Other Side of Digital Wealth. 

The path of digital investment advice is going according to plan. We (with big help from Patrick Davitt @Autonomous) predicted digital wealth to grow to between $500MM-$1.5T in AUM by 2020 in the Fintech Phenomenon analysis, and the latest estimates from Cerulli place is at $220 billion today. Of course most of that is Vanguard, Schwab and other incumbents, which was also expected given the product set and the customer acquisition dynamics in place. But guess what! Roboadvice as a theme is already integrated into the asset management ecosystem and you are too late. So what's next? Well, that depends who you are.

If you are an incumbent, then there is a desperate rush to build artificial intelligence into the investment management product. This is hard, but you can see the investment dollars being poured into the space. For examples, look to Man Group saying to adopt big data or be "eaten alive" (by what? computers or something?), JP Morgan copying BlackRock in creating a quantamental / equity data science unity within its asset management business, and Wells Fargo augmenting research analysts with AI. To see how bankers think about this AI augmentation, see this article on the use of AI at BAML. Augmentation is giving the power of automated human judgment at scale, backed by data, to humans who can apply it on particular fact patterns. More simply, it's letting AI do the first draft, and then having people finish the work.

If you are a consumer-facing fintech startup, then you probably gave up on roboadvice a while back (except for the top 3 or so). Instead, some companies have scaled massively by finding a very concrete paint point and creatin well-designed relief. See SoFi with student loan refinancing, Robinhood with mobile-first stock trading, Acorns with automated savings. In each case, the pain point is immediate and specific -- save $5k on student debt now, buy AAPL without paying $10 now, save $100 this month starting now. But these businesses are too narrow to fill out their current unicorn valuations. So they must broaden. Thus SoFi is going to offer checking accounts (without a banking license, ha!) in the spring. And Robinhood is following neobank Revolut into offering crypto trading on its trading platform.

That makes sense -- compete where traditional finance can't. See how Nordea bank is forbidding employees from trading Bitcoin, or how Vanguard refuses to launch a Bitcoin ETF. We have you on the record Mr. Buckley!

But it doesn't always work out. For example, Stripe is subtracting rather than adding. They were one of the first payments companies to accept Bitcoin payments, but are planning to remove Bitcoin due to slow transaction times and expensive fees. That's been a byproduct of the investment rush, and could be later solved by something like Lightning. But, you know, they say they might use Lumens instead, a crypto coin with the former Stripe CTO on its board. We've already written about rent seeking before, so we'll keep the finger wagging out of this one.

 Source: Autonomous NEXT, Finance Magnates, Bloomberg

Machine Vision Calamities. 

Let's look at how increasing computing power and algorithm efficiency are leading to some pretty wacky technology in the realm of computer vision. The building blocks are as follows. Neural networks can be trained on large data sets of objects to recognize those objects. They run on video cards (GPUs) and power everything from tagging cat photos to Tesla's self-driving cars. The more GPUs, the more things you can recognize, and the better your data and algorithm efficiency, the more accurate your recognition. 

So here's the example -- Amazon and its magic store, Amazon Go. The company has been testing a check-out free shopping experience for a few years, and the acquisition of Whole Foods has only encouraged speculation about the future of food retail. New information has come out about how the technology works. First, a shopper scans an identifier on their phone when entering the store. From that moment on, the hundreds of video cameras on the ceiling watching all the activity in the store track every single shopper and every single product on video. To do this successfully, not only do you need gazillions of hours of footage (i.e., what Amazon is in fact doing), but a massive cloud infrastructure to process the machine vision demands in real time. Good thing there's AWS!

The same neural network that can recognize images can also hallucinate them. Generative neural networks can manufacture images of a type, where the type is their source data set. And if you put an editor on top of that, like an adversary, you can manufacture pretty accurate renditions of whatever if is you want.

Thus, deepfakes. In their current NSFW form (and this is how the trend is being reported), deep fakes use machine vision to swap out the faces of celebrities onto adult entertainment. But that's just the beginning. Using a free desktop app called FakeApp, a derivative of the many mobile face-swap apps, a user can masterfully replace one speaker's face with that of another. And the effects can be good enough to look better than a multi-million dollar 3D rendering by the best Hollywood studios.

Samantha Cole at Motherboard, which broke this article, goes on to say -- "An incredibly easy-to-use application for DIY fake videos—of sex and revenge porn, but also political speeches and whatever else you want—that moves and improves at this pace could have society-changing impacts in the ways we consume media. The combination of powerful, open-source neural network research, our rapidly eroding ability to discern truth from fake news, and the way we spread news through social media has set us up for serious consequences."

Yeah, it's not great. Especially when such messages can be validated for peanuts on social networks using cheap bot armies. According to the New York times, the going rate for 25,000 fairly active Twitter bots is $225. Want to know where the profile descriptions and pictures come from that make these bots look like real users? Stolen identities from humans. 

Source: Devumi retweet sales

Featured

So we are really excited about this!

Lex worked closely with Investopedia to build an awesome crypto currency course that helps make sense of the ecosystem basics, and provides a roadmap for Bitcoin, Ethereum and other top cryptos.

Check out what the course offers here or by clicking on the banner below.

Thanks for reading!

Want even more rich content on #fintech in your mailbox? Our newsletter tracks #blockchain, #roboadvisors, #neobanks, #artificialintelligence, #regtech, #insurtech, #chatbots and more. To subscribe to Autonomous NEXT, click here.

• Follow @autonofintech and @lexsokolin. Readers can also reach Autonomous at [email protected]. 
• Love the newsletter? Let the world know on Twitter!
• This content is not investment research or investment advice, is not intended to lead to a transaction, and should not be used by retail investors (see disclosure).