– When do we trust machines? Are people flawless?

Glenn Henriksen argued in his speech that we expect a higher level of precision from machines and computers than we do from humans. Is this fair, or are we forsaking efficiency for no good reason?

By Lars Johannessen, May 12, 2022

Glenn Henriksen
Glenn Henriksen on stage during his talk on regulations. PHOTO: Stein Johnsen

Head of Technology and System Development at Justify, Glenn Henriksen gave us an interesting look into the use of artificial intelligence in processing legal forms and documents. He set the scene by proclaiming that Justify aim to make law accessible for more people. In terms of being able to understand, but more importantly by making it cheaper. The way to do this, according to Glenn, is by scaling without adding more humans to the tasks, and this is where AI comes in. Software and machine learning have the potential to free up a lot of time that is spent on reviewing standardized forms and documents. There is however a challenge connected to this, and that is that humans can make errors and fill out forms wrongly in an almost infinite number of ways.

Take a look at the images from day 2 at AI+ 2022. Use the arrows to navigate between the pictures. PHOTOS: Sein Johnsen

Read more and see the pictures fraom day 1!

This poses a challenge for AI and machine learning because there can suddenly be an error in a document that is unlike anything the software has seen or trained on. Glenn went on to explain that perhaps this is not the greatest issue after all.

– Ok, so let us say it (the software) discovers an anomaly. We will have it flag that document and send it over to a human who can go through it. The software goes back to processing thousands of documents, and the human resource will go through the few ones that are flagged, Glenn explained.

This led to the point that Glenn was making, we want our software to be without flaw, when we know that humans are not. If we expect ai and machine learned software to be perfect right out the gate, we might be let down. Glenn summarized it by saying “Ai is overestimated in short term, and underestimated long term.”. Ai and machine learning works best over time, with substantial amounts of data to train on, so the best way to do this is to start using them.

Ruth Astrid Sæter
Hostess Ruth Astrid Sæter welcoming the crowd to AI for beginners. PHOTO: Stein Johnsen

Day two of AI+ started with two parallel sessions taking place simultaneously – AI for beginners at Brygga kultursal, and AI for experts at Institute For Energy technology (IFE). Ruth Astrid Sæter, the hostess of this years AI+ conference opened the session at Brygga kultursal and introduced Karl-Magnus Haugen, CEO at airMont to the stage for his talk on how simple, stupid data becomes intelligent with AI.

At the same time in IFE’s auditorium, Division Director for Digital systems at IFE, Tomas Norlander welcomed the audience who joined the AI for experts session welcome and gave a short introduction to his background in AI and IFE’s history in regards to nuclear science and use of AI. Tomas then invited Alexandre de Oliviera e Sousa, Solutions Manager at Cognite AS to take the floor for his talk on AI in industrial context.

Thomas Nordlander
Tomas Nordlander introduces todays speakers for the expert session. PHOTO: Stein Johnsen

The two sessions the continued in parallel until lunch was served. The AI for beginners crowd got to hear Jan Erik Gausdal, senior advisor at Eye-share, talk about “How can AI simplify accounting tasks in your business?”, followed by Associate Professor at Østfold University College, Lars Vidar Magnusson explain AI in “Top-down image analysis”. Elisabeth Haugsbø, Head of Data at HUB Ocean, rounded of the session by explaining how “Data grooming for AI” is done.

The audience at AI for experts was treated to affiliated professor at Arcada University of Applied Sciences, Roberto V. Zicari’s talk on “Trust and AI”. He was followed by Jørgen Torgersen, CTO of Railway Robotics, and Christian Svalesen, Data Scientist at BearingPoint, talk about their project “AI for railway maintenance”. Andreas Risvaag, full-stadd developer at Heimdall Power finished the expert session with his talk on “A startup approach to AI for power grid efficiency” and how AI is used in preventing breaks and issues in the power grid.

Dr. Inga Strümke
Dr. Inga Strümke as she lectures the audience on how machines make decisions without us.

Post lunch was opened by Henrik Fagerholt, Product Manager at Gyldendal Rettsdata, talk about “Lawyers, Law Tech and AI”. Henrik was followed by Dr. Inga Strümke, XAI researcher, TEDx speaker, and Particle Physicist at NTNU deliver an entertaining and somewhat unnerving talk on explainable AI. Inga stated that “Machines might be, and probably are, modelling non-human concepts”, and that self-driving cars have been proven to be mislead by pieces of tape strategically placed on stop signs, making them read it as a speed limit signs instead. Following Inga was Glenn Henriksen, focusing on the possibilities of AI in processing legal documents and regulations for AI.

Torgeir Waterhouse
Torgeir Waterhouse. PHOTO: Stein Johnsen

As the last speaker, Torgeir Andrew Waterhouse, Founder and Partner at Otte, took the stage to deliver his views on “Security and society” and AI. Torgeir gave an enthralling speech on how we could merge democracy, economy, autonomy, and technology into our daily lives. He also pointed out that the need for cyber security has never been greater, and that we as society need to be more aware of this.

– One in four leaders think that they cannot be exposed to cyber-attacks. One in four leaders needs to leave their job, Torgeir proclaimed.

The day ended with a panel debate consisting of Torgeir Waterhouse, Elisabeth Haugsbø, Henrik Fagerholt, and Dr. Inga Strümke. The panel debated issues related to how AI will affect society, how we as humans should relate to AI and machine learning software and answered questions from the audience. A long day of captivating talks, inspiring ideas, and some possible spine-chilling outcomes of AI marked the end of this year’s AI+ conference. We hope to see you again in Halden for next year’s conference!

A special guest visited the opening day of AI+ 2022

By walking across the stage, the robot dog Spot showed artificial intelligence in practice and made everyone in Brygga kultursal stretch their necks as he entered the spotlight at the first day of the AI+ conference.

By Anja Lillerud, May 4, 2022

Spot at AI+ 2022
SPOT IN THE SPOTLIGHT: Annita Fjuke (left), Nils-Ola Widme and Ruth Astrid Sæther were impressed to see how the robot dog Spot moved around on stage. PHOTO: Stein Johnsen

Sitting together with the AI+ 2022 hostess Ruth Astrid Sæther on the stage sofa, Nils-Ola Widme from Abelia and Annita Fjuk from Digital Norway which were the speakers to enter the podium after the opening speech held by Prime Minister Jonas Gahr Støre, were informed that they would receive a special gift from an even more special gift presenter.

From the side of the stage, a blue robot dog entered the spotlight carrying two books on his back. As this is not an everyday sight even for AI enthusiasts, the four-legged robot got everyone’s attention while crossing the scene towards the three people on the sofa.

– Please welcome Spot, Sæther said.

The robot dog is developed by Boston Dynamics and Spot is a platform which Institute of Energy Technology (IFE) uses in research projects. IFE is also one of the initiators to the AI+ conference together with Smart Innovation Norway, eSmart Systems, Halden kommune and Cluster for Applied AI.

Historical soldiers and modern technology

This year’s conference is the third in a row, but it is the first full scale hybrid version with live audience. Halden is of course the place with Brygga kultursal as the conference venue surrounded by the historical city square and the harbor and protected by the Fredriksten fortress on top of the hill southeast of the city center.

A pre-opening of the conference took place on Monday with the social event AI & Beer where associated professor at Østfold University College, Henrik Skaug Sætra, talked about AI for the Sustainable Development Goals.

Historically dressed, the Fredrikshald Borgerbevæpning welcomed the guests standing in the sunshine outside the venue Tuesday morning. Fredrikshald Borgerbevæpning is an organization driven by volunteers aiming to preserve and bring the history of this civilian force which guarded the city from 1644 until 1881, to life.

The historical soldiers were a nice contrast to the subject of the two-day AI+ conference – the modern and rapidly growing technology of applied artificial intelligence.

Aleksander Kostopoulos
SLIDESHOW: Pictures from AI+ 2022 day 1. PHOTO: Stein Johnsen

Wide range of speakers

After a cultural kick off with the percussionist Aleksander Kostopoulos who gave the audience a really wakening performance both visually and audially, the Mayor of Halden Anne-Kari Holm wished everyone a warm welcome to the town, and the Norwegian Prime Minister Jonas Gahr Støre held the opening speech digitally.

Støre spoke warmly about the national AI network and the importance of Norway keeping track of the AI development.

Throughout the day, speakers from Equinor, Digital Norway, Inmeta, Microsoft, Norwegian University of Science and Technology (NTNU), Chronos, Seidr, AI clinic, Oslo university hospital, Kristiania University College, Østfold University College, University of New South Wales in Sydney, Elhub, Abelia and VNNOR held presentations within some of the five topics of this year’s conference:

  • Smart City and mobility
  • AI enabled techonoly
  • Security, ethics and regulations
  • Environment and sustainability
  • Energy

From speeches to dinner

At lunch, the participants had a short walk through the southern part of the town to eat in Haldens Klub, Norway’s oldest eatery of their kind established in October 1786.

The second part of day 1 began with a second performance by percussionist Aleksander Kostopoulos, and 4,5 hours later, when host Ruth Astrid Sæther wrapped the day up and wished everyone a nice evening, members from Fredrikshald Borgervæpning were ready to give a historical guided tour through the city center.

During the afternoon, the conference venue got rearranged turned into a fantastic dining area with great decorations. The first day of AI+ 2022 ended with a pre-drink, a three-course dinner, speeches, and entertainment.

The AI film “Ex Machina” was shown at the Halden cinema theatre after an initiative taken by AI+.

Around 200 people attended the first day physically and about 100 people watched it digitally, and day 2 offers several interesting presentations and new meetings between people.

The future is data-driven: – Data is the gold of any company

One of today’s most long-term and desirable business resource is data. For decades, Norwegian companies have collected and stored data, and with AI as a constantly and rapidly growing technology, we finally have the tools to utilize much of the large amount of stored data.

By Anja Lillerud, May 2, 2022

Jon Jahren på scena
THE VALUE IS IN THE DATA: Jon Jahren is Director Azure Cloud & AI at Microsoft Norway and encourage Norwegian companies to safeguard the real value of their businesses – the data. PHOTOS: Private

Norway and Norwegian industry are in a unique position regarding the scale of data accessible. Only five years ago, the focus both politically and industrially was Norwegian investments in AI and AI research.

Already then Director Azure Cloud & AI Jon Jahren at Microsoft Norway questioned the direction of Norwegian AI development.

– Practically all the algorithms you will need are already open-sourced and the cloud vendors are making them increasingly easy to use, so that’s not where the real value lies, he explains.

Hence, Norwegian companies should concentrate on collecting data and utilize it in the best possible way instead of using resources on developing their own algorithms.

– The data is the gold of any company. Be sure to secure the business gold and let the major tech vendors handle and develop AI, says Jahren.

How to become data-driven

In recent years, more and more companies have been aware of the importance and value of their data, and they have tended to change the use of their data resources and their approach to AI.

At Microsoft’s AI business school, they sum up the journey to become a data-driven organization in three bullet points:

  1. Focus on being a data-driven company
  2. Implement it through an including and joint change of the business’ culture
  3. The use of AI must be responsible and ethical

– The first bullet point is about being aware of the data you collect and how it can be utilized. It’s also a matter of making the necessary tools needed to implement the change available in the organization, explains Jahren.

Jon Jahren

Adopting AI

Another challenge Microsoft faces, is the different industries’ and managers’ fear of the unknown. And even though AI increasingly has become a public domain the last years, it is still unfamiliar to numerous people.

The fear relates not only to one’s own competence, but also to the risk by using AI. Issues regarding ethics and privacy are highlighted, not because of the privacy itself, but because of the AI solution and whether it is secure enough.

One of Microsoft’s major tasks is the work they are doing on developing tools which are using explainable AI. Towards the customers, the algorithms in explainable AI can explain the choices that were made along the way and what is the reasoning behind the result.

Another important task to solve, is what is called differential privacy. In an ideal world, it is possible to build AI models from person sensitive data without the algorithm identifying the data. As an example, Norwegian health organizations have thousands of registers containing valuable data who when being ran through an algorithm can reveal patterns and correlations that humans are unable to see.

– We know there are information hidden in the data that can avert serious illness and improve patient predictions. But as of now, we don’t how to use the data and AI while at the same time ensuring privacy, says Jahren.

A third challenge to handle, is confidential computing. If you need to store or share personal information, you need a secure way to do so. When using a cloud, you must be sure that the cloud supplier under any circumstances doesn’t have an opportunity to access your information. Such solutions are becoming much cheaper and easier to use contributing to an increased adoption of AI into businesses.

– Solving issues like these mentioned, is probably one of the most important contributions from Microsoft, says Jahren.

– Responsible AI is important

Since both trust and reliability are important elements for succeeding a transformation and for adopting AI in an organization, Microsoft have created their own principles and adheres to them to ensure that their AI is responsible.

According to Microsoft, AI must be:

  • Fair and unbiased
  • Reliable and safe
  • Inclusive
  • Transparent
  • Accountable
  • Assuring privacy and security

Because of these principles, Microsoft decides to reject some projects. This is to preserve their integrity. In some cases, the large international company requires a declaration from their AI users to allow them to download the software.

– We cannot control everything and everyone, but by setting clear ethical considerations for usage, we express our opinion and make the user aware of the purpose of the algorithm, says Jahren.

– AI brings new opportunities to mankind

The last decade, AI has become more public, but we are still in the starting pit. The Director Azure Cloud & AI at Microsoft Norway is excited to be a part of the era.

– This new technology brings new opportunities to mankind. The algorithms have become so sophisticated and detailed that they solve issues just as good or even better than humans. Humanity is experiencing a boost by AI enabled applications that I find exciting, says Jon Jahren.

And because of AI being a commonplace, he knows Microsoft will face new and more challenging tasks.

AI is surrounding us and affecting our everyday life, hence there exists a wide range of expectations regarding AI and functionality in the society. The bank customers expect AI solutions to make their bank relation easier and more secure. At home you deal with AI software in your robotic vacuum cleaner and refrigerator.

– There is an anticipation regarding both the functionality of our products and that we offer products solving the tasks that the consumers or businesses need to be solved. This must be done without causing a major competence gap in the market, so it is accessible for a wide audience of developers and users. The days when AI was related only to PhDs in white overcoats are gone, says Jon Jahren.

Join forces to make online life safer for young gamers

Did you know that your sentence structure, use of emojis, speed of keystrokes or choice of words reveals your age group and gender? No? Well, they do. And now is the software AiBA, which manage to identify potential harmful conversations within 20 chat messages, on its way to the market.

By Anja Lillerud, May 1, 2022

FOUNDERS: Professor/inventor Patrick Bours and CEO Hege Tokerud. PHOTO: Nora Kleppe Brennhagen, Oppland Arbeiderblad

Cyber grooming is a major threat to children and youths on the Internet. By impersonating young people through fake online accounts, adults reach out to kids and teenagers on chat and gaming platforms with the intention of making friends with them. Over time they build a trust-based relationship with the children before it turns in to a living nightmare for the young ones.

– The cyber groomers’ purpose is to encourage the kids to send sexualized pictures and nudes of themselves. If they receive any, they demand more. After massive pressure and threats of spread, and in despair to try to avoid the nude photos going viral many young people send new pictures to please the groomer, tells Hege Tokerud, CEO and co-founder of AiBA.

Some relationships escalate because cyber groomers manage to persuade the kids and teenagers to meet in person. Such encounters very often lead to physically sexual assaults.

A proactive solution

By using AI, AiBA trains algorithms to detect potential cyber grooming conversations at an early stage when a connection on a gaming or chat platform where the AiBA solution is implemented, is established. If one is identified, the software alerts the platform owners to make them aware that measures ought to be done.

To ensure the reliability and that correct measures are taken, a human resource is verifying the warning before measures are activated.

– What makes AiBA unique, is the proactive approach to the problem. Our goal is to stop grooming as soon as possible when a relationship to a child is established. Earlier we could detect and identify cyber grooming within 40 chat messages. Now we can do it in less than 20, explains Tokerud.

Expert in behavioural biometrics

Henrik Slettene
CONSULTING: Henrik Slettene, Manager AI and Machine Learning in Inmeta. PHOTO: Bjørn Rosvoll, Inmeta

Behind the technology presented in AiBA, hides the result of years of research on behavioural biometrics. Structure of sentences, use of emojis, speed and rhythm of keystrokes, and choice of words are elements related to age and gender.

Professor and researcher at Norwegian University of Science and Technology (NTNU), Patrick Bours, is an expert in continuous analysis and behavioural biometrics and had the idea of commercialising the research results.

The technological implementation is done by Inmeta, a Norwegian consulting company in Crayon group, specialized in digital and data driven innovation. The company has more than 200 advisors and consultants based in Oslo, Trondheim, Hamar and Bergen:

– When Hege contacted us for a collaboration on developing the technology needed to realize the AiBA project, there was no doubt. We wanted in! This is a kind of project that matters, and Inmeta want to make footprints, says Henrik Slettene, Manager AI and Machine Learning in Inmeta.

Over the last 7 years Inmeta has delivered 150 machine learning projects to more than 50 companies in both private and public sector. With 25 dedicated data scientists and business advisors focusing solely on the Artificial Intelligence domain, Inmeta addresses the complete value chain including strategy development, data platforms, data science and analytical model development, AI operationalization and business change and development.

No legislation in Europe

Access to behavioural biometric data is crucial to be able to train the AiBA algorithm to detect and identify cyber groomers. Paradoxically, lack of data is one of the major challenges the project faces.

– We experience that some gaming enterprises are reluctant to involve in AiBA because they fear to get linked to cyber grooming in a negative way. Still, everyone acknowledges that cyber grooming is a big and growing problem, and we urge the industry to change their approach to the subject, says Tokerud.

In the United States, the industry is regulated by a law obliging the companies to secure their users. A similar legislation has been attempted to be introduced in the EU without success. The industry argue that the legislation will be too resource-intensive to comply with and hence, it will be a dormant legislation.

– But now we have a solution, Tokerud points out.

CHAT LOGS: – Authentic data fram AiBA’s two pilots is important to train the algorithm, says Patrick Bours and Hege Tokerud . PHOTO: Nora Kleppe Brennhagen, Oppland Arbeiderblad

– Your children are safe with us

Two global gaming companies targeting children and youths are pilots in the AiBA project. They are going to implement the software, test it on their game platforms and give valuable feedback to improve the product.

– They think the other way around regarding the company’s name being related to cyber grooming. It is important to them to communicate to parents worldwide that “your children are safe on our platforms”, says Tokerud.

The efforts from the two pilots are crucial to train and improve the AiBA algorithm. In addition, the project also have a collaboration with the Norwegian police to discover if the software can be used to detect cyber grooming conversations in ongoing investigations.

– Developing a software like this is extremely demanding and challenging and a never-ending task. Someone will always try to find a way to fool the system putting us in constantly need of data from authentic chat logs, explains Slettene.

Patrick Bours is attending the AI+ conference together with data scientist Saghar Asadi from Inmeta and they will talk more about the research and technology behind AiBA.