ICT Spring 2019: AI, In depth (Part 2)

ICT Spring 2019 Luxembourg, discussed the latest Digital, FinTech and Space trends and attended by over 5,000 professionals. AI was one of the burning topics – featuring expert discussions about “AI in everyday life”, “Unleashing Human potential” and “Tech all over: From trends to fashion”, and more.

Key AI Trends

Tom Ghelen of Accenture Technology identified 4x AI trends which he believes will become important:

  1. DARQ Power – (distributed ledger technology (DLT), artificial intelligence (AI), extended reality (XR) and quantum computing.). As the business landscape transitions into a combination of digital natives and businesses well into their digital transformations, DARQ is the key that will open unimagined new pathways into the future.
  2. Get to know me – AI will continue to drive the “market of one” ever further into our everyday lives
  3. Human+ Worker – companies need different kinds of worker. Every digital job is generating 4 other jobs, and increasingly HR departments are using technology to select the right people, not only in tracking our digital footprints, but also by testing in augmented reality, and monitoring and ensuring continuity of skills.
  4. MYMARKETS – the idea that companies can hook in to what is currently trending and meet the consumers’ demand for it at the speed of now. An example is that Adidas are now able to 3D print shoes in store to a customer’s unique requirements.

Tony Lee CEO of Saltlux Korea

“What is the difference between human and artificial intelligence? For small data, such as highly accurate face recognition, a human needs to compare 20 photos while an AI system needs about 100,000.”

He went on to discuss embodied cognition, neuroplasticity, collective intelligence and the DARPA “explainable AI” model of deep explanations, interpretable models and model induction. The conclusion is that hybridizing the knowledge models would more or less double the accuracy of AI.

“Einstein famously remarke that the computer is incredibly fast, accurate and stupid. Man is incredibly slow, inaccurate and brilliant. The marriage of the two is a force beyond calculation”.

Who’s going to make money in AI?

Simon Greenman of Best Practice AI talked about an ‘AI goldrush’ in referencing the Californina gold rush of the 19th century. “In the AI gold rush, every inhabitant of planet earth will be generating approximately 1.7 megabytes of data per second, but who is going to extract economic value from it?”

The estimate is that by 2025 the chip manufacturing market will be worth $59 billion, AI and cognitive services $77 billion, enterprise AI solutions (IBM, SAP etc) will be worth tens of billions, and AI derived business value will top $3.9 trillion across all sectors by 2022. At country level, with Taiwanese venture capitalist Kai-Fu Lee having described China as “The Saudi Arabia of data”.

AI – A New Revolution from China

David Duan, CEO, DeepBlue shares that DeepBlue Technology is one of China’s leading AI companies in intelligent driving and vehicle manufacturing, intelligent robotics, AI City, biological intelligence, retail upgrade, intelligent voice, security, chips and education.

He points out that China has been following a deliberate programme aimed at dominating the world AI market, and has largely already succeeded. China currently boasts 26% of the global AI GDP, with the USA second with 14%, and as he points out, because they have no choice, China sees trust levels in AI solutions at an astounding 70%, compared with 17% in the USA and only 12% in Germany.

“This is a big problem for Europe, which has fallen massively behind even the USA, let alone China, in the race for AI dominance, and which means that very soon, Europe will have to accept the role of being a consumer of AI, rather than a developer.”

“From the Chinese point of view, Europe is some kind of a paradise where people already have more or less everything that they want, and that with that comes complacency and risk aversion. The emerging technologies are all about risk taking and there, because China is still institutionally poor, it has a huge advantage because people are willing to risk and fail, and government is supporting that with tens of billions of dollars of investment.”

He cites GDPR as being the death of innovation and questions whether Europe is properly positioned to be an innovator of the future.

In a RoundTable on AI Policies, Cecilia Bonefeld-Dahl said regulation was the key to AI Policies, and that “… AI can avoid bias and is a great tool for risk analysis.” On a remark that while Europe is spending 30 billion dollars on AI, China is spending 300 billion dollars and so how can Europe recover that lost ground? –  Cecilia suggested that it is not even worthwhile to try, the Chinese are already too far away, but that Europe needs to improve and stop having 28 separate markets.

 

Other highlights:

How to build an open AI Nation

David Hogan NVIDIA says that almost every market sector the use of AI is part of core business. From industrial inspection, safety on oil rigs are driven by AI, Smart cities, autonomous cars, biopharmaceuticals, retail and financial – are all increasingly reliant on AI.

He says that the EU invests less than 10% of what China does in AI research so the difference should be in talent, in order to become AI nations.

This can be done through Innovation Platforms (significant computer power is required to allow companies to upscale from proof of concept to production, and this platform needs to be available at an affordable price), Research collaboration (governments cannot make their countries into AI nations on their own, they need to work with researchers and universities and the public sector) and Industry solutions (need to be encouraged and supported by governments

Responding to Fake News

Katrina Fok of Spectee Inc Japan shared that Spectee was founded in 2011 after many rumours in the wake of Fukushima, particularly fake news about foreign looters which greatly hindered the rescue effort. The Spectee platform uses AI based solutions to verify the truthfulness of a story, and they are in partnership with Associated Press and more than 90% of the Japanese news market.

Ethical AI and Fairness Tool

Öztürk Taspinar of Accenture Digital said that AI is driven by data and algorithms, and is already deeply integrated in our lives. “By 2020, 85% of our interactions will be managed by AI.Think about the unintended consequences that AI could bring. For example, Amazon discovered that owing to the skew of historical data, their AI algorithms prefer male employees. In the same way AI credit risk prediction also carries inherent bias, but it is not necessarily the algorithms that are to blame, it is the data. Either way, there is an anthropomorphism of AI, where there is a shifting of the consequences of decisions from human to machine”