Let’s start with a few random ideas

• I want to walk into any store and buy the things I want by using AR to find products I need and Virtual Assistants to help with and personalize my shopping experience without interacting with another human and in a friction-less approach.
• I want a virtual lifestyle assistant that can take care of my mobile plan, insurance, utilities, mortgage, credit and my day to day activity management safely.
• When training online, I want automated, guided learning and a virtual teacher that can adapt to my style of learning. (This can be applied to school systems as well but let’s control our ambitions for now!).
• With medical diagnosis, as a doctor, I want intelligent decision support systems accessible in remote areas so that medical diagnosis isn’t impacted by a person’s proximity to a super specialized hospital.
• We want our national leaders to have access to the apparatus that can forecast threats of any kind with far greater accuracy, and as leaders of business we want access to actionable risk assessment of our enterprise.

All of the above have one common element, the need for increasingly intelligent technology in different forms. When we read fiction of the past, these were things that people dreamed we would have within our reach (including flying cars) by now. So why aren’t these standard fare today except in fantasy media or occasional blips in specialty tech news coverage? Sure there are innovations introduced by technology powerhouses but outside their dominion reliance on intelligent technology is pretty minimal.

The last couple of decades, most of our focus had been on building technology that satisfied our personal and professional needs but tended to stay within a fairly linear approach to developing solutions. We could look at the post Y2K period as an experimental time where many technologies were developed that could be immediately applied to our lives; then a pandemic of global scope validated how important they can be to all of us.

However, as with any successful pilot project, inefficiencies exist, lessons can be learned, and in embracing those insights we can launch ourselves into new innovations with potential to be exponential in their impact.

“Every now and then a man's mind is stretched by a new idea or sensation, and never shrinks back to its former dimensions.” – Oliver Wendell Holmes Sr., Autocrat of the Breakfast Table (1858)

The need at the moment is to focus on improving the intelligence of technology. That objective can’t be served effectively by focusing on the applications we develop; it has to start with changing how we look at the environments or platforms we create to serve our programs.

How is that supposed to be achieved?

A fascinating insight I had recently was in a book called “Sapiens” that talked about the evolution of humans. 70,000 years ago, humans or the version of them that existed then were just another animal group competing with other groups for food and shelter. They evolved from that stage to what we have become now, an Intelligent species due to two key things – Language & Storytelling.

Language was key because unlike other animals that communicated through signals even while using some form of language, humans were able to evolve a broader variation in their communications. The next level of evolution in communication came through story telling or introducing their own variations to observations or in some cases inventing entirely new narratives. How we exchanged information, with the limitations of data and comprehension at that time, and combined it to evolve to create narrations, is probably humankind's greatest invention that has influenced entire populations over the ages. Behind every great creation, whether it is artistic, theological, political, scientific or technological, there is a story that needed to be told that drove entire nations forward. The power of language and storytelling propelled innovation that enabled the growth of the collective intellect of humans over the ages.

Technology is at a similar juncture today as a 70 thousand year old Homo Sapien. However, the part that is missing is efficient communication within what can be considered a peer network of artificial entities. Applications of technology do not have a natural framework or language they can use to communicate with each other whether they are in the same domain or a different one. Knowledge gained with each incremental advancement is contained within isolated ecosystems instead of the growth of the collective intelligence of the technical landscape. These obstacles must be overcome.

A few key points that I believe will allow us to lead the evolution of intelligence in Technology is what I discuss in the following paragraphs starting from basic elements of technology to actions of business leaders.

Great Architecture

One of the reasons for the success of AWS goes back to the foundations of its platform architecture. There is a memo in circulation called “The Jeff Bezos Mandate”, about the initial architectural approach of AWS. Setting aside the origin or authenticity of the note, the contents are spot on in establishing an overarching vision as well as a base framework for building a truly scalable cloud-based hosting platform to include such things as programming and communication guidelines. But make no mistake about it, the clear emphasis was on a platform that served the vision and reserving feature conversations for later.

We no longer should be satisfied with developing minimally viable products operating within the constraints of existing environments. We must extend our thinking toward platform requirements especially in today’s connected world. Consider your own experience where you expect most applications and their data to be accessible from any number of computing devices and any obstacle to that becomes a frustration. The simplest of apps require us to have strong communication, collaboration and productivity focused features geared towards its users, so mediocrity or technical debt in those areas is proving to be more expensive as the needs increase.

This couldn't have any greater importance than in today’s connected world where everyone and their boss has to use technology to function. Technology living within great architecture will not only establish the foundation for intelligent evolution but also will find increasing success with users, while having a much longer shelf-life. Especially, in a world with increasingly low tolerance towards flawed, cumbersome and inefficient applications.

Intelligence Strategy

Something every business today should focus on is a strategy to extract intelligence hidden in their data and using it in the most productive manner. A good number of companies I speak with have this on their roadmap but they struggle in finding a practical, cost-effective way to address it. Companies that do this well and navigate the AI landscape successfully are few and far between.

Should there be a strategy?

Imagine, if asked about their Data Privacy and Information Security policies, a software product owner responds with something like – “We are looking at how Information Security can benefit our business" or “We are experimenting with solutions to implement a secure platform". Thankfully most everyone that owns a software product, understands the importance of security and knows whom to approach to solve this important requirement. Nobody debates the importance of Information Security in software.

The same cannot be said about the importance placed on extracting intelligence from data. There are three things happening at the same time in the Data and AI landscape

• Data production and consumption is exploding at a rate never experienced before.
• Solutions for AI are being invented at a hyper active pace remarkable even for the software industry.
• Businesses are still trying to figure out what they should be doing and how?

When an issue that can be so important to business is experiencing this much change it’s critical that a strategy is crafted to provide a filter through which the company’s leadership can assess it relative to the enterprise’s needs. This can be critical as much to help avoid distraction of valued company resources and time as it is to help surface valuable ideas.

Data Exchange Strategy

Data exchange can help technology owners come together and create a platform for exchanging not just raw data but also intelligence derived within in some form that is useful for another entity. With an exchange in place communication can occur in such a way that the value of the data forms what can become its own story delivering intelligence to its intended audience. Essentially allowing data intelligence to inform in context.

What this essentially means is that there is true democratization of Data and Intelligence to achieve common goals. Going back to my random initial ideas, the AR and the Virtual Assistant do not necessarily need to be made by the same manufacturer, yet they are using the same Data exchange framework to help the human shopper. My virtual lifestyle assistant can talk to a combination of services accessing data that can support its activity management. Similarly my virtual trainer can access the same framework that my assistant is using to assess my workplace performance results to tweak my training agenda. This list can go on and replicate the same real life conditions that help humans in their learning process.

What this essentially means is that there is true democratization of Data and Intelligence to achieve common goals. Going back to my random initial ideas, the AR and the Virtual Assistant do not necessarily need to be made by the same manufacturer, yet they are using the same Data exchange framework to help the human shopper. My virtual lifestyle assistant can talk to a combination of services accessing data that can support its activity management. Similarly my virtual trainer can access the same framework that my assistant is using to assess my workplace performance results to tweak my training agenda. This list can go on and replicate the same real life conditions that help humans in their learning process.

Democratization of Data and AI is not an easy task and is a responsibility that every technology company needs to take seriously. With the knowledge gained in the last few decades about best practices in application development, security, data management and AI, we have the right tools in place to overcome these challenges which may be difficult but not impossible.

Leadership & Governance

A global intelligence framework is probably bound to make some people nervous which is very valid and this is where leadership and governance becomes important, to successfully operationalize this. Leadership being the presence of a motivating vision that sets clear direction not unlike the vision for AWS discussed earlier and Governance being the supplemental element providing clear and consistent rules that keep tactical efforts aligned to that vision without compromising corporate ethics or controls. Without leadership and governance the ideas we talked about until now can become objects in a technologist playground without realistic outcomes.

A Tesla Model 3 is a real life version of a toy RC car. Making the current version of the electric car not much different from a toy car except in scale, was a dream of many but achieving it required incremental knowledge, commitment and vision from the leaders in the space since the 1950s. Leaders therefore have to be at the forefront of the fourth industrial revolution driving the growth of collective intelligence of the technology they are responsible for as well others that can directly benefit from them. Without this, innovations will end up in the IT workshops as POCs in the worst case or localized applications in the best case scenario.

As much potential as AI has there are those that point to doomsday scenarios of AI gone wrong. That seems to be elevated when people combine the potential of AI with insufficient or inefficient oversight. Hence, the importance of combining strong leadership with the important guide rails that governance provides. Leaders can establish governance bodies and frameworks that can establish rules and boundaries within which AI can operate.

A recent research report by SAS titled “How AI changes the Rules", places an emphasis on the new demands AI will place on CEOs, CTOs and CIOs, requiring them to oversee Organizational changes, Information & Data Security and Innovation. Their responsibility needs to extend beyond their organizational boundaries and collaborate with other leaders.

So what rules and goals should the leaders across business functions unite on?

Even leaders need end goals for their objectives and that single point vision when it comes to creating a global framework, should be about creating non-discriminatory machine learning outcomes. Mitigating the risk of unintended outcomes by controlling bias is the fundamental principle of every Data Science use case. Bias is also the key and a very important aspect that can influence outcomes incorrectly or in the worst case negatively for humans, which is why this is such a huge responsibility for leaders.

All of this is easier said than done but is a necessary step and needs the guidance of visionary leaders who embrace technology and create a safe path for technological innovation.

What next?

Technology built on well defined architecture, strategies to address Intelligence and Data and strong leadership can be the tent poles for a global intelligence framework contributing to the evolution of next-gen AI technology. Between those tent poles lie several factors like strong innovation, strong code of Ethics so necessary when the human-technology barriers checkpoints are diminished and a perspective towards contributing to Data & AI as a global property that benefits everyone. An idea that in reality is not so impractical and also not that dissimilar from another revolutionary advancement, the World Wide Web.

References:-

https://api-university.com/blog/the-api-mandate/

https://en.wikipedia.org/wiki/Behavioral_modernity

https://en.wikipedia.org/wiki/History_of_the_electric_vehicle

https://towardsdatascience.com/understanding-and-reducing-bias-in-machine-learning-6565e23900ac