The prospect of artificial intelligence (AI) is both exciting and harrowing. Many industries, such as healthcare, manufacturing, law, and insurance, can benefit by leveraging AI to guide business logic. Still, there are questions as to how to reach the point where the technology can scale to meet demand while avoiding the outcome of a dystopian future where robot overlords rule the streets and humanity struggles to survive.
As it turns out, according to Arushi Srivastava, Senior Director at Japan-based IT firm NTT Data, consensus mechanisms built into blockchain technology, such as Ethereum, can actually prevent AI from running rampant. She told ETHNews, "I think that is where blockchain can really help, because you now have a governance and a consensus mechanism built in the technology." This mechanism prevents a "miscreant bot" from hijacking the network. "So you can build those security measures and blockchain will be able to control such a thing if your AI agent is on the blockchain. That is one good marriage of use-cases that I see between AI and blockchain." She agrees that Ethereum’s blockchain technology can actually stop “Skynet,” so it's more likely that, rather than an AI-controlled bot ordering people around in a seedy biker bar, the bot will be guiding consumers to purchase clothes, boots, and a motorcycle in the near future.
NTT Data has been testing AI with the Ethereum blockchain's executable distributed code contracts for many industries. "One of the proofs of concept that we did last year was about having smart contracts for [internet of things] applications, such as washing machines or microwaves." She said the service provided by these appliances can be enhanced with blockchain technology. When an appliance is delivered, an entry about the customer and appliance is registered onto the Ethereum blockchain. In the event of a part failure or if maintenance is due, an EDCC can take data from an auto-diagnostic program built into the appliance and alert the manufacturer and customer (such as by sending a message to the customer’s smartphone), allowing proper measures to be taken automatically, assuming the appliance is under warranty. Thus, rather than having to wait a long time for a repair facility to conduct diagnostics, a consumer will be able to manage repair issues through their smartphone. Additionally, Srivastava explained that EDCCs can manage payments.
The process of leveraging the blockchain to track repair orders and part failures will generate copious amounts of data. This is where AI can be best deployed. Srivastava explained how AI weaves into the mix to help manage this information. "Because this data is going to be huge you definitely need AI capabilities. This is not a simple business intelligence problem; this requires multiple steps and multiple factor recognitions. That is where the magic of AI can help and that is what we're trying to prove now."
Srivastava still thinks that blockchain technology has to mature further before it can truly allow AI to shine, but she said, "Every day we discover a new problem and solve it." She maintains that the technology needs to be "enterprise-grade" and that distributed data storage also poses an issue. She went on to say, "What we have seen is that scalability is still an issue, and speed of transactions is still an issue. So how you overcome those is the key to actually going to the next step with AI."
AI can make a number of industries, both private and public facing, more efficient, according to Srivastava. She said, "it's a no-brainer" that blockchain-backed smart-grids using internet of things power storage devices leveraging AI could deliver a greater level of efficiency. But she also acknowledges that a number of factors are at play; heavy reliance on EDCCs requires auditing and programmer expertise, skills that many consumers and business users do not have access.
Srivastava also said that AI deployed onto a blockchain structure can lend itself to many applications outside commercial industries like non-profits and NGOs. She said, "I do believe there is a lot of value in getting these technologies so that the whole service angle of non-profits becomes more transparent, and then people would have much more trust for these organizations." Security and value can be realized through transparency according to Srivastava.
Many have recently clamored for AI to be regulated by government bodies, including Elon Musk, who tweeted on the topic.
Nobody likes being regulated, but everything (cars, planes, food, drugs, etc) that's a danger to the public is regulated. AI should be too.— Elon Musk (@elonmusk) August 12, 2017
On AI regulation, Srivastava said, "I do not think that a regulatory framework exists right now in any shape or form." Working in the public sector, it is an issue she deals with on a daily basis. "It's easy to come up with a model, but it's very hard to get it approved and get it agreed to by different people." She sees this issue as an obstacle for AI development. She says this problem mirrors the issues for blockchain technology in the US banking and insurance sector, where the technology is under scrutiny despite acknowledgments of its benefits.
AI technology is very new and the products NTT Data is developing won't be in the hands of mainstream society for a decade or longer, according to Srivastava. She said, in the meantime, it's likely to be delivered to select echelons of society; high-end consumers and enterprise-level applications will see the most benefits from the technology in the next five years. When the whole of society does have access to AI systems, Srivastava said the technology would ease interactions between consumers and manufacturers. "I definitely feel there is a lot of friction being removed from the whole interaction the customer has with the manufacturer or from a dealership."
Srivastava encourages companies to avoid waiting for the technology to be developed, and instead engage in research and development because there are so few players in the space. “If you are a forward thinking organization, I think it is extremely important that you start doing some sort of proof of concept, so that even if you do not right now have an implementable use-case... you at least know where it is right now and what you need to build inside the organization."
NTT Data Services recently announced the formation of a blockchain based consortium with 13 companies in industries ranging from logistics and trade to financial markets, which Srivastava believes is a necessary maneuver for companies trying to develop in the space. She believes that a consortium’s productivity far outpaces the work any one company can do alone – "This is not a one person game. This is best done through consortiums."
ETHNews will provide additional coverage when the consortium officially launches on August 30, 2017.