Presentation is loading. Please wait.

Presentation is loading. Please wait.

Industrializing AI.

Similar presentations


Presentation on theme: "Industrializing AI."— Presentation transcript:

1 Industrializing AI

2 AI is not just for fun anymore ...
… it’s time to demonstrate real results that drive tangible business outcomes

3 Why invest in AI?

4 The AI-data flywheel effect ...
There are a few ways to bootstrap the flywheel:

5 … can create winner-take-all scenarios
But you need to be willing to invest to do AI properly! Yesterday, technology was about enablement. Today, AI is about transformation. So, you need to invest heavily to: Get the flywheel going Rethink the corporate strategy Transform the business to align Implement a data strategy Integrate the AI into your business Manage people change Implement AI-specific governance model and controls

6 State of the market

7 Front-runners are pursuing AI aggressively
88% of leaders increased investments in AI vs 62% of laggards. 1 90% of leaders have an AI strategy and are eager to scale 2 72% of leaders prioritize revenue-generating applications over cost-savings ones 3 Deepening their commitments to AI. Eager to scale AI throughout their enterprise. Prioritize revenue-generating applications over cost-savings ones. -- AI in Business Gets Real, MIT Sloan and BCG, 2018

8 Key innovations have made industrialized AI a reality
Way more compute Way more data New architectures Better simulators Easier to program Faster training and inference More data to learn from Solve more complex problems Unlimited training; speed up time Relatively speaking Deepening their commitments to AI. Eager to scale AI throughout their enterprise. Prioritize revenue-generating applications over cost-savings ones.

9 But there are key factors affecting the pace and extent of adoption
Technical feasibility Cost Labour market dynamics Economic benefits Regulatory and social acceptance Technical feasibility: AI can solve specific tasks, not entire business processes. It’s important to break down problems into component parts to understand feasibility. Cost: technical, transformation, people costs all need to be factored in. Technical - data acquisition, cleansing and integration costs can be significant Transformation - processes need to be changed to fully realize benefits of AI People - retraining Labour market dynamics: with labour arbitrage, sometimes it’s still cheaper to throw people at the problem in certain regions of the world Economic benefits: Top-line growth opportunities make the best business cases Soft factors are critical too - improved accuracy, quality, risk management Regulatory and social acceptance Trust and explainability Controls Cyber security

10 Industrializing AI

11 AI transformation playbook
Build your AI strategy 1 Execute the right pilot projects 2 Build your in-house AI capability 3 Build your AI strategy incrementally!: Design a strategy aligned with the Data-AI flywheel Identify core processes ripe for intelligent automation Analyze, prioritize and bundle opportunities into themes Develop a roadmap of AI assets that create a competitive advantage Build your data asset portfolio Execute the right pilot projects: Identify pilot projects early to gain momentum and experience Key criteria for initial pilot projects: Short in duration (6-12 months) Technically feasible using generally available AI tools Deliver enough return demonstrate value, but limited in scope to contain risk Build your in-house AI capability To complement your AI solution delivery partners, as well as to provide a long-term, sustainable solution Develop a training plan to upskill your AI workforce (not only engineers, but practioners) Develop a recruiting strategy to acquire the right talent Create cross-functional teams of AI practitioners Update your operating model to account for new AI roles

12 A good AI strategy delivers compounding returns
Content Creation Instant global distribution Complexity Content curation 2018: Netflix tied HBO for most wins by a network at 23 each. Netflix got the most nominations with 112 Incremental asset that are added Task / Occupation Business Process Business Model Scope of change

13 Applied AI solutions

14 We are ... World class leaders in Machine Learning and Optimization 1 Cultivate the next generation of AI- driven companies with global reach and impact 2 Training ground for your future AI workforce Business-savvy technologists with deep supply chain expertise 3 We solve your complex business challenges using advanced mathematical techniques.

15 We apply a proven, agile approach to solve complex business challenges using advanced mathematical techniques AI Advisory Blueprint Proof of Concept (PoC) MVP Scale-up Value vs. feasability

16 “AI decodes the nuances of chicken speech”
So, even though AI may not be just for fun anymore … “AI decodes the nuances of chicken speech” Value vs. feasability

17 THANK YOU


Download ppt "Industrializing AI."

Similar presentations


Ads by Google