Contact Us



The future will either be an Internet requiring personal IDs or it won't be
Juanjo Bermúdez - Jan 12, 2016

Rise of the machine

No way. Everything on the Internet is based on servers distinguishing humans from machines. No service can be reliable without that distinction. You can't have a service offering opinions if most opinions are written by bots. You can't acquire new users for your startup by offering a free service if your competitors are using bots to fill forms and saturate your servers.

You can't have a bulletin board service if most messages are spam. And what would our world become without an Internet where you can send messages to other people, express opinions publicly, or build a company from scratch without excessive risk?

We would be back to the nineteenth century. Most of the current economy is based on this enriched interactivity: online advertising, e-commerce. We spend now more time engaged to these interactive services than on almost any other activity in our lives. Now imagine that this stops from one day to another.

Imagine a future where everything has the same level of accuracy distinguishing humans from bots than email spam detectors actually have. Where most messages on a social network are messages from bots pretending to be human to convince you to make a lead for the company they work for. Where you can't trust any opinion because you have no way to distinguish a real opinion from a machine's opinion. That's the future we are facing with the new developments on Artificial Intelligence that will enable bots to mimic almost any human ability.

Privacy vs Operability

“Everything on Internet is based on servers distinguishing humans from machines. No service can be reliable without that distinction.”
All these problems disappear if everyone has a personal ID and they use it to identify themselves on every service. This way, service providers can distinguish humans from bots, and filter data appropriately. This is an obvious solution. So obvious that one has to wonder why it wasn't been implemented since the inception of the Internet. And the answer is privacy. Founders of the Internet had to choose between privacy and operability. There was no known way to guarantee algorithmically both of them at the same time. You had to choose one of them and patch the other. If you offer full privacy (which means anonymity) you have to patch the system against abuses of anonymity; on the other hand, if you guarantee operability (banning anonymous operations) you have to compromise on privacy in some way. In those times, the cost of patching operability was low, and the cost of patching privacy was overwhelming: if you provided full anonymity, you could often distinguish humans from bots using captchas. This could take care of most operational problems of anonymity. Conversely, if you banned anonymity, no architecture was known to provide privacy, which wouldn't involve independent trustee persons verifying it in a costly, inefficient and insecure way. Now we see the consequences of the final compromise they made in our email inbox every day in the form of lots of undesired emails.

Internet voting as a paradigmatic example

A paradigmatic example of these difficulties is Internet voting. If you can develop technology for internet voting that doesn't depend on any trustee supervising privacy, you can expect to find similar solutions to all services on the Internet requiring a compromise between privacy and operability. If you aren't familiar with online voting technologies, probably you are wondering why internet voting hasn't been deployed long time ago in all kinds of elections. It would have the obvious benefits of efficiency and saving costs, and some studies show that it would bring a higher level of engagement from young people to political elections. But, as you know, internet voting isn't being deployed in almost any country in the world. Many security experts argue it's not possible – precisely because you can't provide privacy as well as operability at the same time. If you choose anonymity, you can't distinguish votes from real humans, and if you require an ID, you can't provide privacy on the vote without depending on trusted administrators. So it's a paradigmatic example on how apparently trivial tasks can't satisfy both requirements in an Internet environment.

If we want to save civilization from a not usually considered (but very real) possibility of collapse (yes, it sounds exaggerated, but it's exactly the reality of the situation), we need to start thinking about this problem now, and online voting can be a spearhead on the development of the new technology we urgently need.

I'm proud to be pioneering the work on this field through the Igloovote project, having been able to develop a technology that simultaneously provides privacy and operability in a fully automated way, requiring no trusted parts involved. Amazing times ahead!

[Also published at @Medium and LinkedIn]

Also, read: Why internet voting is a threat to democracy and why it's also the solution