Heikki Ailisto: Combining AI technologies with our traditional strengths
Download as PPTX, PDF2 likes186 views
The document discusses the integration of artificial intelligence (AI) within the context of digitalization, highlighting its various technologies and levels, from narrow AI to super AI. It emphasizes the importance of data availability, computing power, and digital tools in driving the adoption of AI across industries, particularly in Finland. Recommendations for leveraging AI in Finland include encouraging public discourse, fostering innovation, and adapting education to prepare for technological change.
![4419/06/2018
What AI is
AI is a collection of technologies and methods – not a single
technology.
Recently, deep learning neural networks have given great results e.g. for
image recognition
Other technologies and methods include machine learning, natural
language processing, machine vision, robotics, as well as problem
solving & searching, logic & planning & ontologies, probabilistic
reasoning, fuzzy logic
AI Paradox (odd paradox, AI effect):
"AI is whatever hasn't been done yet.“ [Douglas Hofstadter]](https://image.slidesharecdn.com/combiningaitechnologieswithourtraditionalstrengthsailisto-180426102142-180619125531/85/Heikki-Ailisto-Combining-AI-technologies-with-our-traditional-strengths-3-320.jpg)
![7719/06/2018
Industry Believes in Digitalization
A study with 235 major German
companies concluded that “the digital
transition will lead to a significant
transformation … that will require
considerable investment” and more
than 50% of their capital investments
between 2015 – 2019 was going to
digitalization (“Industry 4.0”, German).
[PwC 2015]. This will increase
productivity by 18% in five years and
add revenues 2 – 3% per year.
Drivers:
1) integration of value chain,
2) digital connected products (IoT)
3) emergence of new business models
Global investment in digitalization B$ 900
[PwC]
4800 business executives, managers and
analysts around the world were interviewed
about digitalization. 76% of respondents said
that digital technologies are important to their
organisation’s today, and 92% saw that they
will be important in three years from now.
(MIT Sloan, Deloitte 2015)
Manufacturing (B$ 178), transport (78) and
utilities (69) are the biggest investors in
2016. [IDC]. Respectively the biggest
investment go to operations, freight
monitoring and metering of electricity.
Devices and hw largest now, services grow
faster.](https://image.slidesharecdn.com/combiningaitechnologieswithourtraditionalstrengthsailisto-180426102142-180619125531/85/Heikki-Ailisto-Combining-AI-technologies-with-our-traditional-strengths-6-320.jpg)
Heikki Ailisto: Combining AI technologies with our traditional strengths
- 1. VTT TECHNICAL RESEARCH CENTRE OF FINLAND LTD Combining AI technologies with our traditional strengths Vaikuttajatilaisuus 9.11.2017 Heikki Ailisto Research professor
- 2. 319/06/2018 3 Artificial Intelligence is a part of digitalization trend. Since 2000 we have seen many trends and buzzwords coming. Mobile work enabled by 3G, 4G, 5G Cloud and cloud services Big data IoT, Industrial internet Industrie 4.0 Data Analytics Virtual reality, VR, AR Platform economy AI Machine learning RPA Blockchain
- 3. 4419/06/2018 What AI is AI is a collection of technologies and methods – not a single technology. Recently, deep learning neural networks have given great results e.g. for image recognition Other technologies and methods include machine learning, natural language processing, machine vision, robotics, as well as problem solving & searching, logic & planning & ontologies, probabilistic reasoning, fuzzy logic AI Paradox (odd paradox, AI effect): "AI is whatever hasn't been done yet.“ [Douglas Hofstadter]
- 4. 519/06/2018 5 Levels of AI NARROW, WEAK AI Very good in recognizing faces in images or playing chess or optimizing a complicated system. Totally useless in everything else. STRONG, GENERAL AI AI with broad understanding and human-like consciousness. SUPER AI AI superseding human intellect in all respects. Related to the concept of singularity. Popular in sci-fi. Today, we do not have Artificial Intelligence in the sense that machines would have consciousness or will of their own. (Someday we may have). Maybe we should be talking about “auxiliary intelligence”?
- 5. 6619/06/2018 Why is AI a hot topic now? Data Lot’s of data available at reasonable cost due to internet, digitalization, IoT, SOME, and cheap cloud storage. Computing power Affordable, available and scalable computing power. Hadoop, SPARK GPU, NVIDIA. Algorithms and tools Basically, algorithms were there in 90’ies, but now they have been packaged and made available as toolsets by Google and others. Deep learning NN is great for many apps. Business needs and benefits
- 6. 7719/06/2018 Industry Believes in Digitalization A study with 235 major German companies concluded that “the digital transition will lead to a significant transformation … that will require considerable investment” and more than 50% of their capital investments between 2015 – 2019 was going to digitalization (“Industry 4.0”, German). [PwC 2015]. This will increase productivity by 18% in five years and add revenues 2 – 3% per year. Drivers: 1) integration of value chain, 2) digital connected products (IoT) 3) emergence of new business models Global investment in digitalization B$ 900 [PwC] 4800 business executives, managers and analysts around the world were interviewed about digitalization. 76% of respondents said that digital technologies are important to their organisation’s today, and 92% saw that they will be important in three years from now. (MIT Sloan, Deloitte 2015) Manufacturing (B$ 178), transport (78) and utilities (69) are the biggest investors in 2016. [IDC]. Respectively the biggest investment go to operations, freight monitoring and metering of electricity. Devices and hw largest now, services grow faster.
- 7. 819/06/2018 8 Digitalization and increased productivity usually go hand-in-hand Source: McKinsey Global Institute (case USA) • Five top digitalized sectors have increased productivity by 2.4% every year on average during past 10 years 2005 - 2014 • Respectively, five least digitalized sectors have had no increase in productivity. (McKinsey Global Institute, 2015) • During next five years companies expect to increase annual revenues by an average of 2.9% and reduce costs by an average of 3.6% p.a. (PwC 2016 Global Industry 4.0 Survey, 2000 interviewed)
- 8. 919/06/2018 9 Impacts on industry/ies Adoption of AI is fastest in high-tech / digitalized industries: ICT, telecom Automotive Finance Utilities Media & entertainment Consumer goods Retail Education Health Construction Travel & tourism Source: McKinsey Artificial Intelligence The Next Digital Frontier? Discussion Paper June 2017
- 9. 1019/06/2018 10 So, what are the Finnish strenghts? World-leading in certain sectors: elevators, cruise ships, forest harvesters. Level of education is high on averge (so far). Finns take new technologies in to use faster than other Europeans. Laws, regulation and government sometimes even encourage experimenting and new ways of doing. Research is on par with similar countries. Society with high trust and low hierarchies has an advantage during disruptive times.
- 10. 1119/06/2018 11 Finland is seen as benefiting big from AI trend
- 11. 1219/06/2018 12 Some suggestions for Finnish success Encourage public discussion Find new approaches and ideas: prepare for change start-ups for global growth Means to keep RDI, good jobs, tax revenue and even ownership in Finland should be found Traditional strengths and AI Ecosystems creating new business and value for Finland Regulation and legislation; innovative procurement Good early results, e.g. test area for ships. Health sector could benefit big! ”Learn to learn” Education on all levels must prepare for change. Lifelong learning Embrace the possibilities of AI in companies and public service!
- 12. 1319/06/2018 13 Recommendations for the success of
- 13. 1419/06/2018 14 Quality monitoring tool / Dimecc SIMP Heli Helaakoski, Vesa Kyllönen and Mikko Heiskanen, VTT