Invisible Disruptions : Deep Learning and Blockchain

In the ATOM e-book, we examine how technological disruption can be measured, and how the aggregate disruption ongoing in the world at any given time continues along a smooth, exponentially rising trendline.  Among these, certain disruptions are invisible to most onlookers, because a tangential technology is simultaneously disrupting seemingly unrelated industries from an orthogonal direction.  In that vein, here are two separate lists of industries that are being disrupted, one by Deep Learning and the other by Blockchain.    

13 Industries Using Deep Learning to Innovate. 

20 Industries that Blockchain could Disrupt

Note how many industries are present in both of the above lists, meaning that the sectors have to deal with compound disruptions from more than one direction.  

In addition, we see that sectors where disruption was artificially thwarted due to excessive regulation and government protectionism merely see a sharper disruption, higher up in the edifice.  When the disruption arrives through abstract technologies such as Deep Learning and Blockchain, the incumbents are unlikely to be able to thwart it, due to the source of the disruption being effectively invisible to the untrained eye.  What is understood by very few is that the accelerating rate of adoption/diffusion, depicted in this chart here from Blackrock, is enabled by such orthogonal forces that are not tied to any one product category or even industry.  

Related ATOM Chapters :

Technological Disruption is Pervasive and Deepening

The Overlooked Economics of Technology

 

Comments

[fatcat]

The rate of adoption depends a lot of the perceived utility for the price. The fridges got adopted almost immediately. Or even better the 14-15 game become a craze within a year of it is introduction to the american market. It dissipated quite soon too. All that more than a 100 years ago. My point is that if an invention so extremely useful and affordable it would be adopted in a really short time. I, personally, find blockchain to be overhyped.

[fatcat]

AI, on the other hand , especially GAI, has a tremendous potential. But will be short on delivering for the next few years. For now it is just another buzzword like BIG data and CLOUD. Useful when applied correctly, and yet poorly understood by the majority of the firms

[Kartik Gada]

fatcat,

Cost is certainly a factor in adoption, but remember that any new technology as of 2016's ATOM level will necessarily drop in price quickly.

It took decades for refrigerators and cars to drop in price relative to per-capita GDP, as deflating components are only a small fraction of their composition.

By contrast, anything new (like a VR headset or tablet), can only arrive in an age where everything new it almost entirely comprised of components that decline at Moore's Law-type rates. The fastest-deflating thing of all is software.

Blockchain has a lot of extremely useful aspects, but by the time it is widely adopted, it will be called something else and not noticed by most people.

AI is, of course, the ultimate technology, and the primary basis for the ATOM.

[fatcat]

The peice drop doesn't have to be very fast. For example, the genomes sequencong are stuck around 1k for a couple of years and the utioity of having one is not very hogh for a normal person. If the usefulness of having my genome sequenced was high in improving my health ir lifespan i would pay much more. For the yimebeong even a couple of hundred bucks wouldn't tempt me until there is some tangible benefit i cannot get from family history.

Distributeed Blocchain has one property that makes it fast. it is non-reversability.

This will not stand with the current business practices. And would open the door for a lot of fraud schemes. It is not gonna disrupt the banks that much.

[Kartik Gada]

fatcat,

It does not have to be very fast, but almost any electronics-type technology has a Moore's Law price decline rate.

BTW, the Oculus Rift and HTC Vive, after selling a few hundred thousand units, are now stagnating as early adopters are used up. A price drop will bring in more sales.

Note that the reason low-cost genome sequencing has not disrupted much yet is that it is still only a component of the actual process.

Genome sequencing costs dropping from $1B to $1000 merely means the entire process that would deliver value dropped from $10B to $9.000001B...

[fatcat]

well, even Moore's law seems to be slowing down.. Ray Kurzweil is extremely overenthusiastic about the semiconductor progress. The rality is  nowhere near 6 months between doubling. you estimate about 1.5 to two years. Lately it seems to be 5 years for desktop CPUs and 2.5 y for flash/ssd

[fatcat]

you are absolutely right. Genome sequencing is only one of the many enabling technologies needed to deliver game changing health benefits. ironically they might be extremely useful for GMOs and selective breeding as niche tools . However, it won't deliver anything on its own.

and the whole genome sequencing is stubbornly stuck around 1000 dollars for the last two years. 

d

[fatcat]

does it all mean the progress is slowing down, or there is a huge disruption growing in obscurity only to surprise all of us?

[Kartik Gada]

fatcat,

Going lower than $1000 won't do anything anyway, as the other components of the solution are still way too expensive.