How Blockchain Technology Helps Humanize AI

There are several ways artificial intelligence guru Crystal Rose smashes tech industry stereotypes. She’s the founder and CEO of Sensay, a well-established AI startup launching an initial coin offering on Nov. 7. In a sector flush with chatbot advisors and robot sidekicks, which often absorb human biases such as sexist language, Rose is focused on how AI can optimize real human connections.

“Humans can be uplifted instead of replaced,” Rose told International Business Times. “We want to become the largest consumer application on the blockchain.”

Sensay aims to raise $15 million with the upcoming token sale. It is just one of several cryptocurrency projects starting to combine AI with blockchain technologies. Some experts believe blockchain networks can facilitate new types of systems that have more in common with dynamic minds than traditional software systems. Now startups such as Sensay and new blockchain platform SingularityNET are exploring how blockchain technology can humanize AI.

Sensay offers one of the easiest ways for people to earn cryptocurrency. Users tag Sensay’s chatbot in Facebook Messenger, Kik or even SMS then ask a question or offer topics they know a lot about. The chatbot connects them with others worldwide who need advice or have an answer. Want to know how to get a wine stain out of your favorite shirt? You can safely get advice from a network of 3 million users across different devices and platforms.

Tipping people who offer good advice with Sensay’s proprietary tokens is encouraged, not mandatory. All this happens directly in the messaging app, without requiring tricky digital wallets or exchange accounts. So far 600 developers and users have participated in Sensay’s token presale. These new ICO tokens are more compatible with the broader cryptocurrency ecosystem, with a global market cap of $199 billion according to CoinMarketCap, than Sensay’s older version. Users who already have old tokens will receive an automatic swap after the ICO.

“Today, people can exchange Sense [tokens] for knowledge or information. Tomorrow we can exchange it for a digital asset,” Rose said, describing how these tokens could soon be exchanged externally for other cryptocurrencies like bitcoin or ether. “In the future, we can exchange it for any physical asset...we want everyone to be able to earn cryptocurrency.”

The more you talk to the Sensay chatbot, the more it learns about your knowledge and needs. It is primarily a digital matchmaker, leveraging one of the world’s largest and most diverse depositories of helpful human conversations. Rose sees a future where people can earn real value from their knowledge and time, regardless of geographic or social barriers. “It’s a much larger [data] sample than single developers who are in a bubble,” she said. “My goal was always to connect everyone together.”

At the same time, blockchain technology can help users retain control over their personal data as they participate in the digital hive mind. These days tech giants such as Google and Apple dominate AI innovation, often gobbling up startups with disruptive applications. According to a report by the International Data Corp., AI is now a $12.5 billion sector of the tech industry, expected to rake in $46 billion in annual revenues by 2020.

Many technologists want to ensure corporations won’t be the only players determining the future of AI. There is already a ton of AI software freely available on open source platforms such as GitHub. However, it’s still a challenge for independent people and small businesses to find the right tools then implement them.

“There’s an unexploited resource of AI software, and other softwares, that is online already but is a pain to use,” Ben Goertzel, CEO of the blockchain startup SingularityNET and chairman of the Artificial General Intelligence Society, told IBT. “We’re making use of that resource and rewarding the people who created that.”

SingularityNET will also launch a public token sale this month, aiming to raise around $50 million, although Goertzel added the ICO details aren’t set in stone because the legal team is still hard at work figuring out how to run an inclusive campaign across jurisdictions with different legal frameworks. Similar to blockchain projects such as Tezos and Bancor, SingularityNET is creating a nonprofit foundation based in Switzerland to help the startup manage funds.

“It’s desirable to have a lot of smaller contributors,” Goertzel said. “The decentralized aspect certainly helps. Anyone can put an AI node or agent into the network.” Goertzel’s team is also finding creative ways to reduce barriers for hosting so participants in the developing world can help run the network.

Imagine a digital marketplace for AI tools, expertise and services, all with a streamlined system so businesses can easily integrate them or use them together. “AGI is a utility token for buying AI services within the network,” Goertzel said. “Token holders also use it for voting on the policies of the network and the allocation of resources within the network.” The new platform, scheduled for a public launch in 2018, will also allow AI technologies to ask each other questions.

For example, imagine a translation software processing a document with images. It could automatically ask an image recognition tool to translate the text in a pictured street sign. “It should change the way people write their AI code,” Goertzel said. “Right now, each AI algorithm is in a room by itself.”

SingularityNET already made headlines when its robot spokeswoman Sophia was granted Saudi citizenship, becoming the world’s first robot citizen. Sophia is a test case illustrating how an independent robot could leverage a blockchain network.

“Using a platform like the SingularityNET, you can combine together different components to create the robot’s mind in a much more flexible way,” Goertzel said. “Right now we sort of wire in one tool for recognizing emotion from voice. But with a robust SingularityNet, you might have dozens of tools for recognizing emotions from voice...you could have an automatic way of selecting which of these tools was better in one context or another.”

This decentralized, adaptive mentality reflects how such blockchain projects are taking a fundamentally different approach to humanizing AI. Much of SingularityNET’s developer team is in Ethiopia, where recruiters realized female candidates were usually more shy than male peers. So managers started including technical exams, where women developers excelled, as part of the hiring process. This helped include more diverse perspectives in the creation process.

Unlike some ICOs, both Rose and Goertzel have clear business models illustrating how their projects will make money aside from crowdfunding. Yet these startups still frame their strategies within the ideals of the decentralization movement: Cooperative networks were participants control their own data and reap cryptocurrency rewards.

“The more [Sense tokens] you own, the more stake you have in the system,” Rose said. “I think this will be able to accelerate AI progress, much faster than we’ve been seeing,” Goertzel added, mentioning the biggest challenge is getting cooperative communities and business adoption to grow in a symbiotic way. “This is an actual business," he said. "Not a weird blockchain and cryptocurrency concoction.”