AI and blockchain have a lot in common, and the mistakes that crippled the AI industry in the ’70s are being repeated.
Last week we were part of a deeply insightful discussion with the Managing Director of the Gibraltar Stock Exchange (GSX) and a few of its key members. Gibraltar are introducing regulation for distributed ledgers in early 2018, and with it GSX will become the first regulated exchange for cryptocurrency companies. The room explored the potential for societal change that distributed ledgers provide and the place that our established institutions may hold in the “new world order”.
What was said echoed deeply the more astute discussions that have ranged about the place of AI, and its effects on society.
Artificial Intelligence is a decades-old and fundamentally transformative concept, a collection of technologies tied together more by an ethos than anything else, that are slowly changing the nature of work in the modern world — and with it our society. Similarly blockchain, and the technologies that have surrounded and derived from that concept, offer a potential for transformation — not of the nature of work, but of the nature of stored value and trust.
Where AI has already begun to change the nature of work, blockchain proposes to change the nature of trust.
Whilst that sounds grandiose, Jamie Dimon’s comparison of Bitcoin to tulip mania had some significant merit. That the Bloomberg video covering Dimon’s blockchain speech is book-ended by an advert for IBM’s Watson is perhaps almost too pertinent.
It’s too far to call Bitcoin a fraud or to suggest that it’s going to self-destruct, but there is hysteria there, and there are a lot of people who have vast sums of money invested in it that do not truly understand the technology beneath it. Blockchains are a brilliant technological development with some potentially extremely valuable use cases, but we have seen AI winter before, and we will see blockchain winter before the decade is out if we continue to claim everything can be solved by the blockchain and continue to pump out spurious ICOs.
In the 1970s, AI research was powerful, well-funded and exciting. Huge promises were made about its ability to change the world and solve massive, far-reaching problems, and many of them were correct. Machines can now translate Russian scientific reports into english, fighter jets (and my jukebox) can understand spoken commands and computers can drive cars. However, the timescales in which they were promised were massively underestimated, and many of the technologies did not live up to the hype that surrounded them. The Cold War was long over, and jukeboxes unrecognisable before the Russian reports could be read or the fighter jets redirected.
The resultant backlash from the wider world crushed the industry for well over a decade, cutting funding for AI everywhere and spelling the end for companies that might otherwise have been successful, had the public perception of their visions been tempered with a little more realism.
It may be too late to avoid repeating history to some extent, but perhaps we as technologists can lessen the damage somewhat by toning down the rhetoric around blockchain a little and concentrating at least some of our marketing on its closer, surer goals. This has worked for AI (with some noted exceptions) and consequently the greater goals that were sold in the ’70s are now coming in to reach.
On the other side of the market, investors need to be prudent and look for true value being created, look for iterative use cases that find a place within the existing system, and work to truly understand the technologies they’re investing in, not just the structures built on top of them.
There will be immense value created by blockchain technologies and it will fundamentally change society, but it will take much, much longer than the public is being led to believe, and we will see a lot more casualties before the story is over.
Krzana is the sole real-time news partner of the Gibraltar Stock Exchange.