Preparing for a technological future with Marko Klemetti.
Discussing the future of education, human organisation and AI with Marko in our latest podcast.
Marko Klemetti from Eficode was our guest this week on the AGI podcast for an insightful discussion on the future of education, human organisation, and of course, Artificial Intelligence (AI). Marko is the Chief Technology Officer at Eficode, a company that enables software-driven organisations and builds digital experiences via DevOps, Cloud and UX practices. SingularityNET also collaborates with Eficode and will be hosting a developer workshop in December at DEVOPS 2018.
The future of education and learning
How can we prepare the next generation for the important role that automation will play in the decades to come? This question is central in our discussion with Marko. With warnings on unemployment caused by automation being multiplied, teaching children the tools to stay relevant in the job market and to master impending technological novelties is critical. Yet, Marko takes this assumption some steps further: teaching children about programming, coding, understanding the cloud and IT in general, will allow children to understand their world. The “digital world” that is. Just as physics was gradually included in school curriculums to help children understand their “physical world”.
The first step to that end is to demystify coding. In an attempt to prove that primary schoolers and high schoolers could learn to program working products, Eficode started a program called Code 1 on 1. The first results of that project exceeded expectations. Within 8 weeks, complete novices were able to build complex working products similar to what Eficode had built. While not teaching all the basics to them, the children were logically combining pre-assembled components that they would not necessarily understand in and of themselves and thus create complete products.
Focusing on how children want and can learn
The Finnish education system is quite well in tune to the Montessori type of education where teachers are more “guides” in educations and have a lot of liberty in their teaching which allows them to tailor education to individual cases. As a general structure, however, Marko does contend that it could be greatly beneficial to teach the basics of coding in primary school -maybe with simplified components- and actually producing (self-made) code at a high school level.
This idea of teaching how to code by using simplified “blocks” to be moved around has been used by educational projects such as Scratch, a popular free programming language developed by the Lifelong Kindergarten Group at the MIT Media Lab. While Scratch or Code.org focus on helping “to think creatively, reason systematically, and work collaboratively”, the gamification of such education can also be quite useful to get the child interested. Additionally, Marko points out that combining children’s’ interests in Minecraft or Fortnite, for example, with knowledge on how they work could also prove quite useful. It would be invaluable to teach them how their favorite game works, was built, and could be molded. Finally, the open source culture, whether via GitHub, StackOverflow, or even Youtube are also great tools for self-learning and stimulated learning.
Human organisation: how to optimize human added value.
Echoing Eric Schmit’s take on automation destroying but also creating many jobs and opportunities, Marko reminds us that the inclusion of automation in our jobs has been a long process and often ameliorated the human condition rather than the contrary. Since the industrial revolution, but more so today, machines have made working days shorter, less laborious -at least physically- and more efficient. Yet, we are still far from completely optimizing the few hours of concentration we can dedicate on work tasks, as Mark puts it. While the standard working hours are from 9 to 5, only part of that is usually productive; some will favour focusing on one task per day, others, like Mark, will favour context switching.
Work geographies are also being re-designed as more decentralised models emerge with remote workers and scattered offices being favoured. The model enables individuals to build their work around their life and happiness, and exploit the possibilities of digital interactions at their fullest. The rise of immersive and collaborative VR (virtual reality) and AR (augmented reality) technology is already giving us a preview of the future of remote work. Indeed, the seamless representations we offer our co-workers through our digital identities, be it via Slack or cutting-edge AR, could be enabling healthier work lifestyles in many cases. This model has partly been assumed already by SingularityNET and Eficode alike. Today, the Eficode team is scattered around 6 locations in 4 Northern European countries, and SingularityNET’s team is represented in more than 15 countries!
Similarly, companies such as the Finnish Supercell made incredible gains in the gaming industry by enabling self-sufficient teams to run independently and carry projects from inception to delivery.
While the aforementioned efforts to reform school education in order to incorporate a stronger grasp on technology, are important, similar efforts have to be made in modern companies and with more experienced professionals. Eficode has made it its task to help companies grow efficiently in the information age. If you ask Marko Klemetti: “DevOps” is the name of the game. Indeed, DevOps or Development & Operations entails optimizing “organisational culture, how automation is used for the whole of the organisation’s pipeline and stages of production”. It also includes how AI can be employed to “understand what customers want” and how the organisation can improve on key aspects of their business.
While this may sound like it would only serve some companies, it is important to note that most companies are evolving into tech companies and have software running or underlying some critical parts of their business. Thus, Eficode provides businesses with insight on what technical layers could be explored; how new solutions could be implemented; what strengths within the business could be leveraged; and how quickly they could be ready to adapt to innovative tools. Specifically, Mark emphasised two key terms, namely, DevOps maturity and zero-day delivery.
The first is assessed on a scale from 0 to 100; it requires interviews with key people within the business; an assessment of their expertise and general knowledge; and places the business maturity in contrast with the average maturity identified in its respective industry. Accordingly, proposals for improvements are delivered.
The second consists in defining the readiness of the company in question to adapt to industry-wide innovation. The idea is to reduce the time of adaptation and thus the lag between the company’s ability and their competitor’s ability to adapt to a worthwhile innovation.
What’s next?
The larger question looming over the previous discussion is: how do we organise efficiently around AI and future AI. Robots will not replace all jobs anytime soon and the “human touch”, as Marko calls it, will be relevant for a very long time to come. That human touch will be focused on more intellectual tasks that AI cannot undertake. Nevertheless, not everyone will reap the same benefits from the universal help that AI could and should provide.
One way to look at the partial redistribution of benefits stemming from AI, is the Universal Basic Income (UBI). As Marko contends, the value that is currently being produced solely by humans and that is likely to be produced soon by AI, will not, unlike the jobs associated with it, disappear. In other words, that value will still be generated but not necessarily distributed as to allow those that were not able to adjust to new jobs to “reap the benefits”. UBI could assure that every person in a society gets partly rewarded. Finland has been a relatively successful case study in that respect, although Marko reminds us that the fairly low unemployment rates and a good representation of all socioeconomic groups were definite facilitators.
These types of recalibrations of society according to rapidly evolving technologies are paramount. Currently, the algorithms that run most automated systems have existed for some time, but what enabled us to make proper use of these and AI in general was the increase in computing power in the past two decades. What will and needs to happen next, however, is for new algorithms to emerge and new ways of applying AI to be carved, Marko urged. We need to work our way past the usual classification, recommendation, and pattern identification tasks that we assign to AI today. AI with deeper understanding would be about “iterative thinking”, faster learning, and making AI capable of understanding “when it needs help”. Needless to say that the aforementioned job displacements would take on whole new proportions from such development
Additionally, it is essential to mitigate inherent biases in AI systems as much as possible. This may well come down to learning from the most egalitarian societies. Faithful to his favourite state example, Mark explained that by providing free education and health care to its population, Finland is able to collect information in bulk about its population and more specifically about the socioeconomic composition of its society. In turn, this allowed the country to tailor its healthcare and education models further to its population. The same rationale should be applied to AI. We can not compromise on equal and fair representation of backgrounds when teaching AI systems about the intricacies of our societies.
Finally, Mark foresees AI to become more fluid and individual oriented. Ad personalisation and UI adaptation according to users could result in significant increases in revenues for interested businesses and as user retention strategies. Indeed, it is not hard to imagine a near future where the smart design of an app fluctuates according to our likings and what the AI powering the interface is able to determine about us, and does not just meet a ‘one model fits all’ type of design. Apps like these are also of interest to Singularity Studio.
Afterword
This was a particularly perceptive discussion; filled with great ideas and information that are worth contemplating further. You can listen to the podcast here and also read more from Marko and his work directly on Eficode, Mark’s personal GitHub or his Website.
Other topics discussed and open questions:
- Playable ads
- Neuralink
- Hive minds
- Phone payments
Resources recommended and where to find them:
How can you get involved?
SingularityNET has a passionate and talented community which you can connect with by visiting our Community Forum. Feel free to say hello and to introduce yourself here. Exchange ideas, chat with others and share your knowledge with us and other community members on our forum. We have now launched the #AGICHAT, and we invite you to participate in our themed discussions. Read more about #AGICHAT here.
We are proud of our developers and researchers that are actively publishing their research for the benefit of the community; you can read the research here.
For any additional information, please refer to our roadmaps and subscribe to our newsletter to stay informed about all of our developments.
