In computer science, Moore’s Law dictates that the cost-effectiveness of circuits doubles every two years. While its corollary, Bell’s Law, states that every ten years, a huge drop in the price of processing power causes the development of a brand new computer architecture. It happened when the PC replaced the huge mainframes of IBM in the 70s to 80s, then more recently with Google’s cloud. Are we due for another shift soon?
This article is a brainstorming about the future of big tech companies. It is inspired by Jon Stokes’s article “The Billion User Table” and George Gilder’s “Life after Google” which focus on using blockchain technology to create new versions of the current tech giants.
Why we need alternatives
I think the majority already knows how much we have to trust on big tech companies and the danger that comes with. But just to point a few out, here are some risks:
- Filter bubbles
- Privacy issues
- Power consolidation
- Single point of failure
That being said,when should tech companies be regulated?
Jon suggests that we should look at the size of a company’s users table. Software lives or dies with the size of their users table. That’s because it offers you access to network effects, so
- the more users your platform has, the more value it has to any one user (cf. Metcalfe’s law)
- it’s easier to attract new users
- it’s harder for any one user to leave (because it means to give up the value)
If you want to cap the size of a platform, limit the size of the users table. Moreover, it’s already an issue that everybody has separate users table, which are only accessible through an API. Jon gives a good example for all the non-technical people:
“If I have endorsements for JavaScript and WordPress, and I register with your new website via a LinkedIn button, your site isn’t automatically going to know that I have these endorsements. In order for me to learn about your LinkedIn endorsements, I have to hit a LinkedIn-hosted API to get that information.”
This architecture has some drawbacks. Taking this example, LinkedIn is in control of the users proprietary endorsement data and has therefore leverage over it. This lets the data unprotected against any LinkedIn specific business risks. We have no other choice but to hope nothing bad happens.
How are we going to solve that?
The Blockchain as the Internet’s users table
Jon lays out the possibility how the blockchain can be used as a public users table at Internet scale.
Taking a blockchain as the users table, tech companies would have to build their applications on top of it. So the identity-hosting blockchain (e.g. BitClout) represents decentralisation at the datastore implementation layer, and recentralisation at the datastore access layer.
This comes in handy for all users, because:
- if anything happens to an application, the public would still have access to the old data
- we could add various skills under our own identity like LinkedIn, Github, … and decide which to recognise/ignore
Startups will strongly profit from this implementation as well. If everybody shares the same users table, all e.g. Github users are now LinkedIn users by default. Therefore, every new company has immediate access to network effects as described earlier.
Since everybody has already an account, companies would try to convince users from using their application actively.
Perhaps this would lead to spam. In order to prevent that, Jon suspects that there will be paid inboxes. Either way, I think if we use a blockchain as a public users table, it will lead to:
- more freedom of speech (obviously)
- more security
- more competition between companies, therefore users get better services and products
- new ways to get an income (third-world countries will profit a lot)
- accelerate the process to a borderless world
A public users table would dethrone the whole advertising/data collection giants like Google and Facebook. But what about other business models?
A shift in hardware manufacture
As I written in the Introduction, Bell’s Law states that every ten years, a huge drop in the price of processing power causes the development of a brand new computer architecture.
I’m taking Nvidia as an example. In the past, they went after fast, but energy-demanding “hot” chips. Now they are moving to cooler and more efficient ones, enabled by the cheaper circuitry now available. These types of processors are far more efficient in running multiple functions at once. Especially in emerging fields such as sensors for self-driving cars, this would be crucial.
However Google and other information giants are still stuck on the task of achieving cheaper and more powerful processing power to support their massive and growing centralised servers.
This change will also assist the rise of decentralised computing power. Now that chips constantly becoming cheaper and faster, we can get rid of centralised cloud services.
There are already promising projects, for example Golem. Built on the Ethereum blockain, Golem describes itself as the “Airbnb for computers”. You can rent your computer’s processing power when you’re not using it. Users can now perform difficult tasks much cheaper, and without requiring the massive and expensive hubs of Siren Servers like Google’s.
How will companies react to this change?
On the public users table:
At the beginning, only startups/small companies will embrace it, because big corporations don’t want to lose their economic moat. But with a growing users table, it will absorb companies continuously. Network effects suddenly work against them, so they have two choices:
- Join the blockchain
- Get left behind
However, I think even some big companies are looking forward for this shift. Twitter is a good example:
On the shared processing power:
This is more difficult to say, but I think that shift will be even faster. Due to the direct comparable price for computing power, people will choose the cheapest. Therefore, Google and other centralised cloud providers will go down with their prices. But people will still tend to use decentralised system, because of security/privacy reasons. Finally, these tech companies will shut their services down.
How can we accelerate the process?
We need to convince as many people as possible to start using decentralised systems. I think the best way to do that is through education, at least that’s why I started. A lot of people know the danger of big tech, but just a few are aware of the alternatives. But once we convinced enough, network effects will take care of itself.
It’s like with everything, the beginning is always the hardest.
I just hope the majority switch to decentralised systems by themselves, before they are forced to. In the past years, controversial actions and lawsuits by big tech companies increased exponentially and it doesn’t seem to stop.