2 The appliedai initiative is rooted in the UnternehmerTUM ecosystem...
The business plans of the next 10,000 startups are easy to forecast: Take X and add AI. This is a big deal, and now it s here. Kevin Kelly, Founding Executive Director, Wired Magazine
6 Why is this a problem?
It's not just the new capabilities of machines VS General AI Narrow AI 7
It s the type of those capabilities 1 2 3 Computer Vision Computer Audition Linguistics Machines do now have these capabilities for the first time cheaply available. 4 5 6 Mathematical / Logical Machines Interpersonally Intelligent Machines Motion & Robotics 8
It s the speed of development 9 Source: NIVIDIA (Left), Skydio (Right) URL (Left) : https://www.youtube.com/watch?v=clvcgemkjs0 URL (right): https://www.youtube.com/watch?time_continue=87&v=gh5pat1o2v8
It s the falling cost-factor Student research group Roboy Student startup Ampelligence 10 Source: https://www.fastcompany.com/90234055/the-hunt-for-red-tide-relies-upon-ai-and-retirees
Compute plays a major role: [This technology will drive] the third wave of silicon processors Ty Garibay - Forbes Technology Council 11
Source: Google Source: Google IoT Robotics Positive feedback loop AI AR Source: OpenAI
Technical perspective on AI Economical perspective on AI 13
This all leads to: Cheap prediction PREDICTION is the process of filling in missing information. Prediction takes information you have, often called data, and uses it to generate information you don't have. - Prediction Machines: The Simple Economics of Artificial Intelligence by Ajay Agrawal, Avi Goldfarb, Joshua Gans Framing a problem in terms of an prediction problem is massively helpful Data Item Dimension 1 Dimension 2 Dimension 3 Data Item to predict Dimension 1 Dimension 2 Dimension 3 Dimension 4 14 Source: Google A.I. Experiments: Visualizing High-Dimensional Space
An economical perspective on AI Extensive demand due to massive price drops Cheap prediction Accuracy matters! But it can t do... - No one cares Substitute due to capability and price Rising value of its complement data Rapid price-fall of associated products 15
Only 23% of companies we surveyed have actually deployed AI. David Kiron, MIT SMR (Dec 2017) McKinsey MGI: 20% (June 2017)
AI will allow to form competitive advantages - But first central challenges need to be tackled 1 Positive overall view on AI 2 3 4 Strategic questions need to be addressed New roles in companies and the engineering disciplines New application-areas are supercharged 5 Cooperation is the key to success 17
1 Positive overall view on AI 18 https://www.slideshare.net/thebostonconsultinggroup/artificial-intelligence-have-no-fear?linkid=53313274
AI Adoption Cultural change Collaboration Knowhow & Talent AI Project setups Data, Dev, Digi & AI 2 Strategic questions need to be addressed What is your AI vision? Growth via Products Efficiency via Processes Aligned with targets (Customer Experience, Reduced Complexity/Risk, Increased Speed) Vision Defining frame and scope Prioritization based on strategy AI Use Cases Organizatio n People AI execution Blueprints 19
3 New roles in companies and the engineering disciplines AI Engineer Grow Engineers and Developers towards AI Software Engineers AI Project Lead A unique role that acts as ambassador and leader for AI. AI Strategist Develop your strategic view on AI and build up your competitive edge. Employee in the Age of AI Understand what AI will enable and change in your everyday life. 20
4 New application-areas are supercharged Wirtschaftswoche 21
5 Cooperation is key to success Handelsblatt 22
23 Lifting Germany to the AI Age: The appliedai Initiative serves as unique, central point of contact to guide Germany and Europe to the AI age
Since 2018 the strong network of the initiative is growing Strategic Technology Partners Public Partners Strategic Industry / Network Partners Academic Partners IT-Solution Partners 24
With our product portfolio, we tackle the key challenges Germany and Europe face while transitioning to the age of AI Three key challenges identified in UnternehmerTUM-AI study: Knowledge Exchange Tailored offerings of appliedai Difficult translation from AI to business Lack of AI knowhow und talent Proof of Concept Challenging implementation Education Context 25
Knowledge Exchange: We foster the exchange of knowledge and experience in AI within our partner network Industry and officials Students & Startups Tech scene and thought leaders Decider and government 26
Proof of Concept: we support the implementation of ideas with our development expertise and technology infrastructure Intelligent solar panel sensors Autonomous ship Autonomous indoor drone Project Hackathon based RnD Experiments (aai calls this PoC or TMS stage) Hackathon based RnD... Hackathon based RnD 27
28 Education: we help our partners to develop AI skills internally by educating talents within the workforce
Since 2018 the strong network of the initiative is growing Strategic Technology Partners Public Partners Strategic Industry / Network Partners Academic Partners IT-Solution Partners 29