Blockchain and AI are two of the most discussed technologies in the last ten years, but until recently both have been developed independently of the other. The idea of combining them has been around for years, but remained in the realm of speculation.
This is due to the enormous market for the two technologies. Gartner, a global technology research firm, has estimated that the business value created by AI will near $3.9 trillion in 2022, while some anticipate the blockchain market will be worth roughly $23 billion by 2023.
Another driving force is that both AI and Blockchain have seen a significant increase of adoption in recent years.
Research undertaken by the consulting firm Mckinsey & Company shows that 47% of businesses are already using AI technologies in some way, and 78% are planning to increase their investment in this area.
With regard to Blockchain, a PwC survey has found that almost 84% of businesses has been actively involved with the technology.
Clearly, combining the two technologies represents a huge opportunity. Many companies are already actively working with one or both of AI and Blockchain, and have critically invested in the technologies.
But are they really compatible?
Given the level of investment in both AI and Blockchain, it was inevitable that opportunities for combining them have been proposed for years. However, it has not been immediately implemented due to the doubt on whether they can work well together or even at all.
In some ways, AI and Blockchain work in the same environment. They have both been developed in response to the huge increase in the number of networked users, systems, and devices over the past few years.
On the other hand, each technology has (so far) been used to achieve different outcomes. AI, for instance, has been typically deployed to analyse the vast amount of data generated by highly networked systems.
Blockchain has been developed for a different purpose: to secure exchanges of information between anonymous partners.
The distributed ledger that is an integral part of blockchains ensures that a huge network of individual users can securely agree on the contents of a chain, but in some ways is conflicting to the way in which AI works.
AI typically requires that a large number of data is stored in one place for analysis: in blockchains, this storage is decentralized.
The applications of AI and blockchain
There are three key areas in which combining AI and Blockchain have been explored.
The first has been to leverage the extra security that Blockchain provides in order to improve the security of AI.
AI works best when it is fed with huge numbers of data points for analysis, but these data sets are vulnerable to hacking, theft, and tampering.
Blockchains, on the other hand, potentially offer the capacity to store huge amounts of data securely. In this sense, the two technologies may be able to complement each other.
Another speculative application of the two technologies is that it could increase the computing power of AI.
Large AI systems often require huge amounts of computing resources, whether in terms of CPUs or simply data storage. It may be possible, therefore, to use blockchains to automatically identify and utilise distributed computing resources for AIs to use.
One of the most exciting possibilities here is that some have suggested that AI machine learning could significantly reduce the power consumption and requirements needed for mining cryptocurrency.
Third, it might also be possible to combine the two technologies and a third one: the IoT. Using blockchains to secure IoT networks is already being implemented, but as these devices proliferate it could be that AIs can be used to manage them.
All that said, there remains challenges for any system that aims to unite AI and Blockchain.
The primary concern in combining the two technologies has been security. Managing IoT devices through public blockchains, for instance, raises huge concerns about the vulnerability of personal data to theft and surveillance, especially given the rise of cryptojacking.
Another issue is that the speed at which AI processes data is simply not matched by most blockchains. Ethereum, for instance, remains one of the largest blockchains, but can only process roughly 15 transactions per second. AI’s require much quicker access than this.
These challenges will be difficult to overcome, but are not impossible. At the same time, AI is also being used to improve cybersecurity.
Similarly, while there remain concerns about the security of Blockchain, in many applications the technology has greatly increased the security of data.
Both technologies, according to Ludovic Rembert, Director of Research, at PrivacyCanada.net, bring unprecedented privacy risks and also opportunities to improve data security.
Some successes, a few of which are detailed below, have appeared on the horizon, and we should expect more to follow, along with a spectacular implosion or two along the way.
Though applications that combine AI and Blockchain remain rare and experimental, there have been some success stories already appearing.
One example is the recent initiative by Cortex, who have announced the launch of an AI-based network for decentralized applications, designed to improve and optimize financial services.
The idea here is that AIs will be used to generate credit reports for users on decentralized ledgers, and can also be used to detect fraud and other scams.
At the development level, Singularity Net have focused their attention on trying to overcome some of the compatibility issues that affect any attempt to use AI and blockchain together.
They offer a decentralized network, itself based on blockchain, that aims to make the development and monetizing of AIs easier.
Back in the real world, there are also some examples of consumer-facing apps that use a combination of AI and blockchain. Cashaa, for instance, is using a combination of blockchain and AI to enable P2P lending, and Taipei is using blockchain, IoT, and AI technologies in its bid to become a ‘smart city’.
As with any developing technology, it is difficult to predict what the specific applications of AI and Blockchain will be.
One thing is clear, though: both technologies will continue to develop at a rapid pace. Many countries have launched initiatives that aim to encourage developers to explore the possibilities of both technologies, including USA, UAE, and even Malta.
The outcome of these programs remains to be seen, but in order to predict the direction of travel perhaps we should return to where we began: the purpose for which AI and Blockchain were developed.
AI, in short, was developed to analyse large amounts of data, and make predictions based on it. Blockchain, at its core, was made to ensure the security of large data sets.
Though making blockchain-based data sets available to AIs still presents many challenges, it remains as the most logical way to combine these technologies.
If it can be done, the security of AIs will be greatly improved, and the vast amounts of data held in blockchains will be available for analysis.
Editor’s note: e27 publishes relevant guest contributions from the community. Share your honest opinions and expert knowledge by submitting your content here.
Image Credit: Markus Spiske