James Manyika at AI Frontiers: Sizing up the promise of AI
1 like1,057 views
1) Artificial intelligence is progressing rapidly, with machines now matching and exceeding human performance on tasks like image recognition. 2) There has been significant investment in AI from technology companies, venture capital firms, and private equity firms in recent years. China and the United States have received the majority of AI investments. 3) While some industries and early adopters have begun implementing AI technologies, adoption remains limited even among firms aware of AI's potential. Barriers include a lack of technical capabilities and uncertainty around return on investment.
James Manyika at AI Frontiers: Sizing up the promise of AI
- 1. November 2017 Artificial Intelligence The next digital frontier? JAMES MANYIKA CONFIDENTIAL AND PROPRIETARY Any use of this material without specific permission of McKinsey & Company is strictly prohibited Rated #1 Think Tank 2016 (private sector category) by the Global Think Tank Index, University of Pennsylvania
- 2. 2McKinsey & Company Machines can now match and even exceed human performance Image recognition accuracy rate 2McKinsey & Company Present2010 70% 80% 90% 100% Human performance Machine performance
- 3. 3McKinsey & Company 3McKinsey & Company Who is investing and where?1 Who is adopting AI today and how?2 Where can AI improve business performance?3 What do we need to do tomorrow morning?4 Four key questions for AI today
- 4. 4McKinsey & Company Investments in AI include several types of technologies Computer vision Natural language Machine learning Virtual assistants Robotics and autonomous vehicles 28% 56% 7% 6%4% EXTERNAL INVESTMENTS
- 5. 5McKinsey & Company There is increasing interest from VC and PE funds Electric vehicles +285% Biotech IT Total VC Energy AI AI investments in 2016 VC investment growth, 2013â16 External investments (VC, PE ) $6â 9B
- 6. 6McKinsey & Company Tech giants are the primary investors AI investments in 2016 External investments (VC, PE ) $6â 9B $20â 30B Corporate investments (tech giants) 42 19 1413 99 7 2010 20122011 20162013 2014 2015 Number of AI acquisitions
- 7. 7McKinsey & Company China and the United States dominate the investment scene, while Europe is lagging North America $15â23B Europe $3â4B Asia $8â12B AI investment in 2016 External investments (VC, PE, M&A) Internal corporate investments
- 8. McKinsey & Company Higher potential Media Consumer Energy Agriculture Manufacturing Public/socialHealth careAutomotive Finance TelecomPharmaceuticals 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Impact score Travel, transport, and logistics Volume Identify fraudulent transactions Personalize financial products Identify and navigate roads Personalize advertising Optimize pricing and scheduling in real time Predict personalized health outcomes Optimize merchandising strategy Discover new consumer trends Personalize crops to individual conditions Predictive maintenance (manufacturing) Diagnose diseases Optimize clinical trials Predictive maintenance (manufacturing) Lower priority
- 9. 9McKinsey & Company Yet adoption at scale is limited, even among AI-aware firms 20% early adopters % 40 Contemplators 40 Experimenters 2 technologies 1 technology 3+ technologies 7 3 10 9McKinsey & Company
- 10. 10McKinsey & Company What about the non-adopters? 28% donât have the technical capabilities to implement AI effectively Technical capability Return on investment 41% of AI-aware firms are uncertain about AIâs ROI 30% of AI-aware firms canât see a business case for AI Business case 10McKinsey & Company
- 11. 11McKinsey & Company Sectors leading in AI adoption today look set to pull away further from slower adopters Future AI demand % â AI spending 2017â20 Current AI adoption % of firms who are early adopters Slower adopters Frontier sectors 5 0 10 15 25 3010 2015 Automotive Tech and telco Finance Construction CPG Transport Retail Media Health care Energy Education Professional services Travel 11McKinsey & Company
- 12. 12McKinsey & Company Digitally mature industries lead in AI adoption AI Index Digitization Transformation Index 80 40 20 100 0 60 60 40 80 Travel Construction Finance Retail Tech and telco Transport Health care Media Professional services Energy Education 12McKinsey & Company
- 13. 13McKinsey & Company The early adopters of AI have five distinctive characteristics 13McKinsey & Company Digitally mature1 Larger businesses2 Adopt AI in core activities3 Focus on growth over savings4 C-suite level support for AI5
- 14. 14McKinsey & Company AI can more than double the impact of other analytics across industries Public and social sector 100% Travel, transport, and logistics Telecom, media, and technology Pharma and medical products Healthcare systems and services Global energy and materials Advanced manufac- turing industries Consumer FIG Additional AI impact Analytics impact ~60% ~40% PRELIMINARY SOURCE: McKinsey analysis
- 15. 15McKinsey & Company AI use cases have the potential to materially impact corporate performance PRELIMINARY SOURCE: McKinsey analysis Potential impact as % of revenue Advanc- ed elec- tronics aero PMPTele- com InsLogis- tics Media Auto Basic mate- rials Oil and gas PSSHSS Trans- port TravelCPGHigh tech Chem and agri- culture Retail Banking Average ~10%
- 16. 16McKinsey & Company AI use cases have the potential to materially impact corporate performance Advanc- ed elec- tronics aero PMPTele- com InsLogis- tics Media Auto Basic mate- rials Oil and gas PSSHSS Trans- port TravelCPGHigh tech Chem and agri- culture Retail Banking Average ~10% SOURCE: McKinsey analysis PRELIMINARY High data availability, particularly labelled and structured data Potential impact as % of revenue
- 17. 17McKinsey & Company AI use cases have the potential to materially impact corporate performance Advanc- ed elec- tronics aero PMPTele- com InsLogis- tics Media Auto Basic mate- rials Oil and gas PSSHSS Trans- port TravelCPGHigh tech Chem and agri- culture Retail Banking Average ~10% SOURCE: McKinsey analysis PRELIMINARY Operationally oriented manufacturers can realize significant improvements in supply chain and production Potential impact as % of revenue
- 18. 18McKinsey & Company AI use cases have the potential to materially impact corporate performance Advanc- ed elec- tronics aero PMPTele- com InsLogis- tics Media Auto Basic mate- rials Oil and gas PSSHSS Trans- port TravelCPGHigh tech Chem and agri- culture Retail Banking Average ~10% SOURCE: McKinsey analysis PRELIMINARY Pharma and med products companies can unlock additional R&D productivity Potential impact as % of revenue
- 19. 19McKinsey & Company AI use cases have the potential to materially impact corporate performance Advanc- ed elec- tronics aero PMPTele- com InsLogis- tics Media Auto Basic mate- rials Oil and gas PSSHSS Trans- port TravelCPGHigh tech Chem and agri- culture Retail Banking Average ~10% SOURCE: McKinsey analysis PRELIMINARY Industries with significant physical plant and equipment assets can optimize their capital expenditures Potential impact as % of revenue
- 20. 20McKinsey & Company Potential AI impact by function varies by industry Travel, transport, and logistics Telecom, media, and technology Public and social sector Healthcare systems and services FIGConsumerGlobal energy, and materials Advanced industries 100% Pharmaceu- ticals and medical products Organization OtherOther operations RiskStrategy and corporate finance Service operationsManufacturing Marketing and salesSupply chain management Product development PRELIMINARY SOURCE: McKinsey analysis
- 21. 21McKinsey & Company Techniques and tools Open organization Workflow integration Data ecosystem Use cases/ sources of value 21McKinsey & Company There are five pragmatic steps businesses can take to start adopting AI tomorrow
- 22. 22McKinsey & Company Governments need to act now to promote AI development 22McKinsey & Company RegulationSkills Support and incentives
- 23. Download MGI research at www.mckinsey.com/ @McKinsey_MGI McKinseyGlobalInstitute
- 24. 24McKinsey & Company Colors: MGI 2017 0 0 0 Dark 1 255 255 255 Light 1 51 51 51 Dark 2 254 254 254 Light 2 197 197 197 Accent 1 51 189 242 Accent 2 52 133 198 Accent 3 51 52 51 Accent 4 242 127 0 Accent 5 128 128 128 Accent 6 181 179 140 Hyperlink 0 105 131 Followed Hyperlink 0 154 166 Teal 162 173 0 Apple 255 208 0 Yellow 205 32 44 Red 181 95 0 Rust 102 48 124 Purple 11 121 163 Ocean 26 66 99 Dark ocean 125 154 170 Blue-gray 221 221 221 Light gray Default shape Fill: Accent 1; No line Default line Accent 1, 3/4 point Main palette MGI custom
- 25. 25McKinsey & Company Colors: MGI 2017 for thinkcell Light grayOceanPurpleRust Blue-grayTeal Dark oceanRedYellowApple Apple Yellow Ocean Light gray Red Purple Blue-gray Dark ocean Rust Teal