Balaji S. Srinivasan is the CEO of 21.co, a start-up that’s proposing a system where strangers can pay investors and other busy people to respond to their emails, as well as a board partner at Silicon Valley venture capital firm Andreessen Horowitz. He has a Ph.D. in electrical engineering from Stanford and has been an early investor in crypto-related projects and currencies including bitcoin, ethereum, and Polychain.

He’s written about cryptocurrencies and tokens frequently, including this piece 3 months ago about the power of tokens.

After causing a stir in July when bitcoin hit $3,000, the digital currency is now at $5,000.

Here’s an edited version of our conversation from last week (and audio of our discussion is located here on my podcast):

Eric Jackson: You can trade private shares on a grey market today, what is so special about the blockchain for facilitating a market for private assets?

Balaji S. Srinivasan: So, first of all, I just want to say that in my view the equity analogy is incomplete. Especially if doing a blockchain or token in the US, to be compliant with all regulations you need it to have a utility component and not just use it as a proxy for equities. This is not too hard to do, as there are so many interesting things you can do with tokenized networks. So while I won’t be a stickler in every point in this interview for making that distinction, it’s a very important one to keep in mind: you want to work with your lawyers to ensure that tokens or blockchain-based assets in general aren’t equities.

With that said, relative to existing secondary markets in either property (like houses) or equities, the scale of what is happening with the token market is unparalleled. It’s one thing to have a bunch of small secondary markets; it’s quite another to turn the internet itself into a market. Anyone with an internet connection can purchase a token. See shapeshift.io for a great example of this.

The rise of all these new tokens and public blockchains means the internet will, in the long term, become by far the biggest “stock” market — once the regulatory issues are worked through — just as it has become the biggest library.

Jackson: How do you see the regulatory aspect of tokens playing out in the coming years?

Srinivasan: The first point is that it’s very important to remain compliant with all available regulations in your jurisdiction. From a US perspective, the good thing is that the SEC has made clear that there are some forms of tokens that are not securities. For example, they referred to ethereum itself as a currency, and they have granted that legitimate innovation in the space is possible.

The second point is that this is an international phenomenon; like bitcoin regulation, we’ll probably see many different paradigms for token regulation around the world.

Jackson: You often hear that, if the SEC doesn’t embrace cryptos, the community will move to places like Zug, Switzerland. What’s your view?

Srinivisan: It’s not really a future event — the community is already there and growing fast. Crypto is already far more geographically decentralized than (say) search or social. The major players in many other tech verticals are often in the Bay Area and Seattle; the major players in crypto are already spread all over the world. Ethereum for example has its major two loci in Brooklyn and Berlin, with its foundation in Switzerland.

Jackson: Why won’t the SEC require tokens only to be sold to accredited investors?

Srinivasan: Without speaking for them, I believe their position as outlined in their filing is that if it doesn’t pass the three prongs of the Howey test then it may not be a security. Obviously not everything that appreciates in value is a security. For example, a house can appreciate in value, but you can (and many people do) buy it for the use value. A domain name likewise can appreciate in value, but it’s not a security. And from another standpoint a token is similar in many ways to an API key, but that’s also not a security. It would be crippling to the US economy to restrict the sale of houses, domains, or API keys solely to the 3% of Americans that are accredited US investors.

Jackson: What can and can’t be tokenized?

Srinivasan: Tokenization applies to scarce assets. Today the most appropriate thing to tokenize is something that’s purely digital. Bitcoin and ethereum are the canonical. Filecoin and Tezos are promising. I think as of 2017 it is somewhat fraught to try to tokenize something like a private asset or a stock given the regulatory uncertainty.

With that said, by say 2025-2030 I expect that there will be multiple jurisdictions that allow the tokenization of virtually any scarce resource, all the way down to personal tokens. That is, like Upstart.com, you may eventually be able to take a stake in an individual’s tokens, giving them a lump sum of digital currency today in exchange for a smart-contract-enforced percentage of their future earnings.

Jackson: What’s the upper bound for all tokens? Some point to $9 trillion in value of gold holdings, some point to use cases like increased value of unlocked private value – like a bump of 10% in the value of all REITs – others point to inherent value of transactions (TPV)? How do you assess what different tokens are valuable?

Srinivasan: Anything scarce will ultimately be tokenized, because the benefits of digitization and increased liquidity are so great. That means cash, stocks, bonds, commodities, houses, cars, digital goods of every kind, and perhaps human time as well in the form of the personal token described above.

The regulatory framework will likely eventually accommodate this. Only a few countries need to allow it, and the consequent creation of wealth will be so large that it’ll push many of the rest to have a liberal tokenization regime as well.

Jackson: Someone tweeted this week that ICOs this year were like the glut of fiber optics in the dot com era; it facilitates later build out of the space. Is this a fair analogy?

Srinivasan: Yeah, it’s the Carlota Perez / Gartner Hype Cycle thing all over again. Virtually every major technology has an initial spike of interest, then a dip, and then a long-term rise to success. The dotcom bubble is the canonical example, but there are many more. bitcoin alone has gone through at least 4 of these cycles.

Jackson: If the current craze is like the dot com era, are we in 1996 or 1999?

Srinivasan: Hard to call the timing as that’s like calling the top. Certainly you can make an ethereum 2017/Netscape 1995 comparison, or a bitcoin 2013/AOL 1993 comparison. Each wave of hype gets a bunch of smart people working on important problems, and that continues even after the hype recedes.

Jackson: Is tokenization going to lead to a barbell approach to VC where some firms become super-angels and others become more like hedge funds?

Srinivasan: I think tokenization eventually means everyone becomes an investor, once all the regulatory issues are worked out — from your computer itself to a kid in India messing around with $10. It also means everyone from angel to VC to growth to public markets investors starts to buy the same assets. Kind of like how packet switching turned many disparate forms of information (photos, video, audio, text) into packets, the blockchain is turning many disparate forms of scarcity into tokens.

Jackson: There are a lot of crypto hedge funds cropping up in the past two months. How do you assess their quality?

Srinivasan: Look at the team behind them, and of course their track record.

Jackson: Are we headed to a world where every business has a token?

Srinivasan: Eventually, probably yes. If you believe that recent Bloomberg article, frequent flyer miles are evidently a more profitable business for airlines than actually flying people around.

Jackson: What are the cryptocurrencies that are most interesting to you (beyond bitcoin and ethereum)?

Srinivasan: ZCash, Tezos, Filecoin

Jackson: Are all forks in cryptocurrencies a form of hidden inflation?

Srinivasan: I think they are more like a completely new way to consensually avoid inflation, and run actual experiments. If we could have forked the dollar in 2008 and had an exchange rate between BailoutDollars and AusterityDollars, probably AusterityDollars would have traded higher. But we don’t know because we weren’t able to run the experimental comparison between an inflating and non-inflating economy.

Jackson: How will we know which crypto companies today are going to grow to become the Amazons of the future?

Srinivasan: They’ll say then what they always do: the big ones had great founders and teams. But that’s usually only obvious in retrospect, and a necessary but not sufficient condition. Many great founders have one or more big failures on their track record. What makes them great is that they eventually succeed despite that.

Commentary by Eric Jackson, sign up for Eric’s monthly Tech & Media Email. You can follow Eric on Twitter @ericjackson .