Next Steps In The Integration Of Artificial Intelligence And The Blockchain

While a lot of AI technologies are owned and operated by centralized providers, a majority of the blockchain players in the market publish all of their …


Having worked in the cryptography space for over two decades, and having been an active participant in the cryptocurrency evolution since its inception, I take a deep interest in the subject.In particular, I believe that the intersection of artificial intelligence (AI) and blockchain is an exciting but challenging new development.

Matt Turck recently discussed why the topic matters and highlighted interesting projects in the space, referring to AI (big data, data science, machine learning) and blockchain (decentralized infrastructure) as the defining technologies of the next decade. Evidently, the time is already ripe for these new concepts, despite them being novel and still underdeveloped.

Intriguingly, AI and blockchain are philosophically different in various ways:

  1. AI is driven by more centralized infrastructures as opposed to blockchain’s decentralized, distributed nature.
  2. While a lot of AI technologies are owned and operated by centralized providers, a majority of the blockchain players in the market publish all of their codebases as open-source code that is freely available for anyone to inspect at any point in time.
  3. AI is more of a black-box solution for now, while the blockchain tends to be more transparent in all the transactions processed.
  4. AI is based on probabilistic formulas, while blockchain is more deterministic in nature.

Currently, AI startups are being overwhelmingly acquired by companies such as IBM, Apple, Facebook, Amazon, Google, Intel and Alibaba, among others. These organizations rely on unprecedented amounts of data to train their AI agents, which offers them an immense competitive advantage. At the same time, their data and capabilities are closed from the rest of the world.

Unfortunately, centralized AI introduces room for abuse, such as massive surveillance of people using face recognition and computer-vision-powered technology. At the same time, creating solutions on top of a centralized environment requires enterprises to give up privacy and control of their data to other third parties.

Merging AI And Blockchain

This is where blockchain comes in, as it can be used to overcome many of AI’s shortcomings. I’ve seen it firsthand in our business, where we leverage a lot of AI and machine learning (ML) capabilities in order to better identify and authenticate users’ blockchain identities.

Currently, experts in this space are exploring ways through which blockchain can be deployed to create a decentralized marketplace to enhance AI. This course by MIT is just one indicator of the movement in this space. This will allow people to comfortably share their personally identifiable information (PII) with the assurance that it will remain secure and private through decentralization and secure computing offered by the blockchain. In effect, users can easily share their sensitive details, such as health and financial data, and the system would ensure that only the intended service provider would have the ability to decipher and decrypt users’ PII with explicit consent by the user. With time, I believe the space will have an accumulation of massive data maintained by big organizations using AI algorithms to stay competitive.

An article on Hackernoon lists some of the latest projects integrating blockchain and AI technologies to create cutting-edge solutions. Some of the more notable ones include SingularityNET, an AI marketplace where enterprises can acquire AI capabilities on a global scale to enhance the growth of the space. Another project is Namahe AI, a platform that aims to improve the efficiency of supply chains by integrating AI and blockchain to enable seamless monitoring of the processes in real time and flagging anomalies and fraud for review. Finally, there’s Numerai, an AI-based hedge fund that sponsors competitions for industry enthusiasts to develop and submit prediction models and solutions.

Challenges Of Merging AI And Blockchain

Obviously, AI solutions differ from legacy ones, since they follow probabilistic models. In other words, a traditional program follows the approach of “IF A happens, THEN follow B.” On the contrary, AI (deep learning and machine learning) uses probabilistic answers to follow a succeeding step. This feature of AI makes the technology ideal for creating flexible solutions. Nevertheless, the tradeoff is that some AI programs make mistakes.

To date, AI agents still go wrong in some cases, and it still remains difficult for users to know when it is wrong or what should be done when it makes a mistake. A few memorable examples include a Microsoft chatbot gone rogue, Wikipedia edit bots engaging in feuds among themselves, Uber’s self-driving cars ignoring red lights and Russian robot Promobot IR77 escaping the laboratory.

Another issue is compliance. It is still a major concern to control AI solutions from going rogue or causing damage. AI and blockchain solutions will require data aggregation, which is a real challenge. However, the internet of things (IoT) will be vital in the provision of data required for AI training. In effect, the security and privacy of privately owned data will be crucial in this space.

Talent is another challenge for merging blockchain and AI. While data, which is the primary factor for training AI models, can be gathered using IoT devices, professionals will be needed to develop algorithms that run in a decentralized or distributed manner as required in blockchain technology. Fortunately, organizations such as Deep Brain Chain and SingularityNET are continuously researching and creating innovative AI algorithms.

Computing resources still remain an issue in merging AI and blockchain. Luckily, it is possible to leverage global idle computing power to run resource-intensive AI training integrated with blockchain.


Some experts are now suggesting blockchain has the capability of decentralizing AI to achieve decentralized intelligence available to the masses. To reap the real benefits from their integration, I believe it will be imperative to address several major concerns: how to determine when an AI solution is wrong in its operations, how to train professionals in the field and the need to come up with appropriate compliance requirements to guide the development and deployment of the products. To make real progress, today’s players in the field should seek to break down these roadblocks and encourage blockchain and AI development in the real world.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?

Related Posts:

  • No Related Posts

Richard Branson gives Elon Musk some advice: Learn to delegate and get some sleep

From one business visionary to another, Virgin Group’s Richard Branson has given embattled Tesla founder Elon Musk some advice after his recent …

Like Branson, who started life as a student entrepreneur and then founded the Virgin Group, which controls more than 400 companies, Elon Musk is seen as something of a visionary in the business world.

His creative approach to business and technology has seen him launch pioneering projects such as his electric car brand Tesla. Musk also dreamed up the idea of an ultra-high speed transportation system called Hyperloop, as well as a space tourism business called SpaceX. Branson and Virgin are direct competitors of Musk, among other players in the same field, as they are also developing hyperloop and space tourism programs.

Unlike Branson, however, Musk is a controversial figure, particularly because of his increasingly erratic behavior and tweets — ranging from accusations of pedophilia against a British cave diver, to stating that he was going to take Tesla private, much to the shock of shareholders.

Needless to say, Musk is facing legal action for the former, and has received a $20 million fine from the Securities and Exchange Commission for the latter.

Related Posts:

  • No Related Posts

Artificial intelligence startup Machinify secures $10 million in Series A funding to develop first “Data …

That is the problem the Palo Alto, California artificial intelligence startup is trying to solve. Machinify is a new artificial intelligence startup that applies …

Did you know that 2.5 quintillion bytes of data are created every day? Organizations around the world face the challenges of making sense of the data and getting the right insights needed to make informed decisions. That is the problem the Palo Alto, California artificial intelligence startup is trying to solve. Machinify is a new artificial intelligence startup that applies the latest advances in cognitive computing to analytics.

The company announced today it has raised $10 million in Series A funding to develop first “Data-To-Cash” AI platform. The round was led by Battery Ventures, with participation from GV and Matrix Partners. Machinify was created to give every enterprise access to AI technology that can enhance its core business and unlock significant untapped value.

Founded in 2016 by former executives from VUDU, a digital-video company acquired by Walmart in 2010, Machinify offers an easy-to-use platform to any company that seeks to extract the maximum value from its data. The team’s VUDU experience led them to see the great potential for AI to improve business performance by automating decision-making, as well as the great operational challenges involved in the process.

Machinify is led by founder and CEO Ganesan (the former CTO of VUDU), fellow VUDU alumni Tony Miranz (former president), and Edward Lichty (former GM). Machinify’s founding Executive Chairman is Alain Rossmann, who was previously a founder or co-founder of VUDU, PSS Systems, and OpenWave. The company is backed by AI experts at Battery Ventures, GV, and Matrix Partners. Machinify aims to dramatically reduce the cost, time, and complexity of designing, putting into production, and managing sophisticated AI-driven business decisions. The company utilizes AI technology to help businesses digest their data and unlock hidden profits through automated decision-making.

“To date, few companies are able to surmount the operational hurdles involved in deploying AI decision models to improve their business performance, and those that have made an attempt often fail to meet their schedule or to achieve the real-world ROI they planned for,” said Prasanna Ganesan, Machinify’s founder and CEO. “We are completely reimagining the process of developing, deploying, and maintaining AI-powered solutions. As a result, business owners at Fortune 1000 companies are able to leverage Machinify to transform data into cash.”

“The Machinify team is a rare find that blends an academic understanding of AI’s potential with a history of commercial startup success, and we believe they have created a transformational product,” said Max Schireson, a Battery entrepreneur-in-residence who works closely with Machinify. Schireson was formerly the CEO of MongoDB (Nasdaq: MDB), an open-source database company.

Added Dharmesh Thakker, a Battery general partner active in AI investments: “Machinify arms non-technical domain experts with the power of AI. For example, healthcare businesses are using Machinify today to significantly increase revenue and profitability.”

“Machinify is laser-focused on the critical operational issues facing the deployment of AI-driven software within enterprises,” said Adam Ghobarah, general partner at GV. “This new generation of software is dynamically driven by AI models and large enterprise datasets. It requires a completely different approach, and we believe that the Machinify team and platform can help enterprises unlock more value.”

Related Posts:

  • No Related Posts

Is Artificial Intelligence Smarter Than Humans?

In this opinion piece, Clay Media Consulting creative consultant Clay Morrison (pictured below) tackles artificial intelligence’s actual intelligence.

In this opinion piece, Clay Media Consulting creative consultant Clay Morrison (pictured below) tackles artificial intelligence’s actual intelligence.


This is just one of the questions in everyone’s minds as AI becomes more and more commonplace in everyday life. What can AI do that humans can’t?

And what can humans do that AI can’t? A new infographic, Rise of the Machines, is tackling these comparisons head-on.

The key difference with AI is the ability to learn – this term refers to software which manifests some kind of intelligence.

It might be man-made, but it can operate in a way that’s based on how the human brain works. That’s where it gets a bit confusing.

How do we demonstrate, test or measure intelligence?

The aim of AI is to learn, adapt and think.

It must demonstrate behaviours we associate with intelligence. That can include any combination of the following:

  • Planning
  • Problem-solving
  • Reasoning
  • Learning
  • Perception
  • Recognition

It could be easy to dwell on how much further AI has to go, but that would undervalue how much it is impacting us.

Here’s what the technology is already capable of doing:

  1. Speeding up processes – Machines can do things a lot quicker than us humans, including boring monotonous tasks we’d rather not do ourselves. They can also do it with a lot more accuracy.
  2. Improving security – In the right hands, machines aren’t influenced by external pressures. They can’t be bribed or blackmailed and they can be alert to risks at all times. The core purpose of some AIs is to hunt down and identity threats. For example, they are often used by banks to detect fraud.
  3. Remembering stuff we tend to forget – AI doesn’t get distracted and let something slip its mind. It’ll remember to change the heating setting, to remind you that you owe someone money or to recommend films based on what you’ve watched before.

However, the biggest argument against AI so far is that it doesn’t have the same emotional responses as humans.

While we are able to use compassion and empathy to inform our decisions, AI is programmed to function a specific way.

However, whereas narrow AI sticks to a rigid set of tasks, general AI is designed to use previous information to adapt its responses, so although it’s not quite human, it can still learn from experiences the way we would.

The lack of human traits can also be a positive thing. AI doesn’t experience boredom when completing monotonous tasks, which means it’s able to carry them out with more accuracy at a quicker pace.

It’s also impervious to outside influences like bribery or blackmail, so it’s an excellent type of technology to use for security.

Could AI replace us?

You aren’t alone if you have hesitations about the growing involvement of technology in our lives. Change is daunting.

If films have taught us anything, it’s to fear how robots could destroy us. But that’s far from reality – for the time being at least.

“By far, the greatest danger of Artificial Intelligence is that people conclude too early that they understand it” – Eliezer Yudkowsky

What’s more, we’re at risk of jeopardizing future developments if we exaggerate the risk.

It’s a point Bishop Director of Microsoft Research in Cambridge, emphases.

“Any scenario in which [AI] is an existential threat to humanity is not just around the corner,” Bishop told the Guardian.

“I think they must be talking decades away for those comments to many any sense. Right now, we are in control of that technology and we can make lots of choices about the paths that we follow.”

“I believe this artificial intelligence is going to be our partner. If we misuse it, it will be a risk. If we use it right, it can be our partner.”

The future of AI lies in closer cooperation between humans and machines. We certainly look forward to seeing how we use AI to enhance – not threaten – human possibilities.

Sign in with LinkedIn

Please login with linkedin to comment

AI artificial intelligence

Related Posts:

  • No Related Posts

Book Highlights MindMeld: CEO and Artificial Intelligence – Afterward Future Thoughts

Even with the explosive pace of technology, human endeavors in artificial intelligence are still very primitive. The pace of the pursuit of human-made …


By Thomas B. Cross @techtionary

The following is an excerpt to Afterward: Future Thoughts of MindMeld: CEO & AI Merging of Mental & Metal book available now from iBooks – Available on iPhone, iPad, iPod touch, and Mac.

Book Review – “As the CEO of a energy industrial company and actively involved in CEO Leadership Forums I have been following AI for more than a decade. Indeed the promises for improving many technical tasks are interesting yet in reality often prove more complex to manage than proposed. MindMeld was very profound in proposing that AI starts not at the bottom of the organization but with CXO decision-making and worth reading by anyone in or rising to the boardroom.” George B.

For interviews, professional guidance, product/market research or evaluations, articles, speeches or presentations as well as CEO Executive Seminar on AI, please contact

Here are some of highlights from Future Thoughts, please click on any image for ibook:

There are still many key issues requiring further research. Even with the explosive pace of technology, human endeavors in artificial intelligence are still very primitive. The pace of the pursuit of human-made systems remains, for the most part, an extension of machines rather than an extension of human knowledge processing. If this approach continues, there will be vastly sophisticated high-speed processing devices that will provide elegant simulations of real-world conditions. However, these conditions will parallel the real world rather than interact with it. The development of new technology continues to outstrip society’s ability to absorb its effects, much less its potential consequences. However, technology is like the extension of a rubber band stretched far ahead. While society is anchored in the past and holds the rubber band back, technological change stretches the rubber band ahead. The rubber band keeps getting stretched further until, sometime, it will break, splintering society into widely diverse factions. Everything from terrorism to AIDS is being impacted by the increasing rate of communication and the knowledge of underlying issues. With the acceleration of technology, the concept of artificial intelligence becomes an even more perplexing enigma. There have been business problems that have always existed: labor, working conditions, inventory, supply, and finance, and now there are issues, such as environmental impact, career planning, virtual everything, and productivity. Other new problems are on the horizon, and these merge with the old problems. Together with the needs of competitive advantage, new market demands, and global distribution, networks provide little opportunity for human-made systems to be up-to-the-moment. This doesn’t make the study of artificial intelligence worthless. AI permits organizations to exist where none were possible before. Much as one 100-horsepower motor took the place of 100 workers in textile mills a century ago, AI will become the intelligent worker of the future. Some of the major technologies that offer the greatest potential are:

1 – Neural networks. The subtle but truly amazing ability of the mind, upon a second’s glance, to recognize a face unseen for ten years and, at the same moment, forget where the car keys are. Automated neural networks offer nearly all industries and humankind new intelligent tools that become increasingly more human with use. This is partially due to the fact that most business is based on human interaction and human thinking processes. Systems that begin to learn about humans in humanlike ways, rather than humans adapting to machinelike ways, offer an incredible business opportunity.

2 – Visualization systems. Visualization systems are those systems that convey and process information graphically. Rather than machines that see, these are complex machines that are able to potentially understand doodles, notes, ideas, and images as humans do. Humans interact with one another and their world in nearly a totally sensory way, and for the most part, in a totally visual way. Images are formed, sequences are organized, events are cataloged, and. life-spans are archived, frame after frame. Visualization systems process information and provide, like pages in a book, frames or windows of events, concepts, and emotional situations. With advances in image processing and storage technology, within a few years people will be able to start recording important events in their lives at a phenomenally low cost. The ability to record every moment is less critical in itself than the ability to use this information as part of a person’s expertise or skill, not unlike the skill of a successful CEO or manager whose ability and related compensation come from their knowledge, manipulation, and organization of certain facts to the financial benefit of an organization. By capturing personal information and by organizing it in a certain way, individuals will be able to market their own automated data bases to organizations, much like “human” software programs. Throughout history, the power of information derives from, among other things, its ability to move from one point to another—its portability—and its usability when it arrives. Employment potential has also meant being in the right place with the right skill at the right time. The greatest problem with the demise of the industrial era is that in an information society skills can rarely be transmitted from one generation to the next. New skills are required, and people’s ability to organize information and themselves is critical to their overall success. The development of systems and software that allow people to develop their own personal data bases of information and provide that as marketable information will be as important in the near future as the latest generation of hybrid was to a farmer one hundred years ago. Visualization systems are not robot vision devices that can navigate through a maze; they are systems that interact with humans and with the world in a visual way. More than intelligent graphics, these systems convey knowledge and interact with other visual systems via visual language, operating in much the same way as a speaker who might begin a presentation by saying, “Let me tell you a story.”

3 – Idiot systems (dumb expert systems). The problem that presently exists with most expert systems is that they either require so much expertise that it takes an expert to use one, or they solve problems within too narrow a scope to useful in the real world. What is really needed is an ignorant machine or idiot system (IS) machine. An IS knows nothing but is willing to learn about anything; it’s a system that does not pretend to be the expert, but makes a good apprentice. This system can tolerate enormous amounts of ambiguity, logic jumps, and gut reactions. It’s more like a pencil than a smart pad.


Related Posts:

  • No Related Posts