An Overview of Applied Artificial Intelligence
Artificial intelligence is the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings. On this page, you’ll find a deeper explanation of applied artificial intelligence (applied AI) as well as links to additional resources.
Artificial intelligence is rapidly moving from the laboratory and into business and consumer applications. The result is a fundamental shift in how software is built, and what it's capable of doing. And while we're still a way off from the artificial general intelligence portrayed in the movies, artificial narrow intelligence is a reality that's already powering some of the most successful technology businesses today, including Amazon, Facebook, Google and Apple.
Principles of Applied AI White Paper
Taking advantage of applied AI as part of your business solution isn't easy and so to help we have written this paper that introduces the 10 principles of applied AI, and provides framework and maturity model for getting started and avoiding common mistakes
What Is Applied Artificial Intelligence?
Applied AI is the use of artificial intelligence to enhance and extend software applications.
To accelerate the use of AI in the software companies we invest in, Georgian Partners has developed a pragmatic framework to assist the adoption of machine learning and other building blocks of AI: The Principles of Applied AI.
Our investment thesis is that businesses that adopt applied AI will outperform the market because AI will enable superior service levels in terms of capability, delivery, availability, accuracy, convenience, engagement, cost and brand.
The Best Applied AI Companies Today:
- Use machine learning to automate decision-making with high accuracy.
- Put a high emphasis on continuous learning from process and feedback data.
- Use human experts-in-the-loop to fill in the gaps.
- Proactively monitor for quality and bias.
The Principles of Applied Artificial Intelligence
There are similarities between applied AI and applied analytics (AA). That's because applied analytics can be an example of AI models applied to the delivery of insights into a business process. However, applied AI extends the applied analytics framework with:
- An increased focus on end-to-end process automation.
- Use of advanced machine learning techniques.
- Better articulated performance objectives and model accuracy.
- Appropriate use of human judgement and automated predictions.
- Continuous learning and adaptation through feedback loops.
- Approaches to limit the impact of model error and bias.
The 10 Principles of Applied AI can can be divided up using the following framework:
Start with the Processes
AI has the potential to affect every business process and job role either through augmentation or automation. Business value can be impacted by injecting insights into existing processes, optimizing processes, or automating entire or parts of business processes. Beyond that, AI provides opportunities to redefine or create entirely new processes that weren’t possible before, even with human effort.
To identify the places with the greatest potential for AI in your business, start with the processes your business controls or should control, and where improved outcomes or efficiency would have the highest benefit to your and/or your customers’ businesses. Understand which parts of the process involve human judgment, and which of those parts should be optimized, augmented, automated or reinvented.
Understand all processes.
Prioritize the most valuable processes.
Integrate artificial intelligence-enabled processes into your business via APIs for software and intuitive interfaces for people. Integrate human judgment as required to improve predictions. Leverage your customer base, in-house domain experts or crowdsourcing engines to supplement model predictions. Enable continuous learning by observing, measuring and improving outcomes as actions are delivered within a business process.
Design frictionless integrations.
Integrate human judgment as required.
The first step in model selection is to determine the desired level of performance from a business perspective. Select techniques to build machine learning models that achieve desired outcomes and to fit budget, objectives and available training data for each of the most valuable predictions and integration points. A lack of planning and achievement of objectives for error and bias will increase business risk and reputational harm. Business differentiation will come from the optimal combination of model integration, the modeling techniques themselves, and the data used to train and optimize the models.
Understand performance objectives.
Start with proven modeling techniques.
Machine learning is being deployed everywhere, and it’s fueling today’s artificial intelligence revolution. In this episode, Jon Prial talks with Parinaz Sobhani, Georgian Partners’ new…
The type and volume of data required for applied AI depends on the machine learning technique(s) selected. This data may be that which you already collect and have access to (albeit perhaps with additional labeling), or it may need to be acquired from your own sensors or from third parties. Develop plans and infrastructure to capture and cleanse relevant data to support your AI models, and adjust frequency and breadth as necessary.
Capture relevant data to support models.
Monitor performance and fairness, and build customer trust. Ensure sufficient understanding of models and data to satisfy expectations of repeatability and interpretability. Adopt a market leadership position on ethical, privacy and legal concerns related to use of artificial intelligence.
Manage quality for error and bias.
Build fault tolerance as part of model integration.
Defend your legal and ethical stance.
Adopting the principles of applied AI will be an evolutionary process for any organization. The starting point for every organization will vary, although all can work toward higher levels of adoption of each of the principles over time. The best companies will hire smart, curious, capable machine learning scientists and engineers, and allow them to keep current on emerging machine learning capabilities.
All the principles are important for the creation of the best AI solutions. However, don't look at them as a playbook for implementation. Read through them all, and become familiar with the intent of each. Your own implementation will likely start with understanding the processes, but you will iterate in your thinking about integration, models and data. Managing quality and bias, and dealing with governance issues, is something to be mindful of throughout the entire process.
Machine learning has become so ubiquitous that it’s behind many of the technologies we use every day. And yet few people really understand it. That’s…
A lot of effort goes into identifying who we are almost from the very moment we’re born. Birth certificates, passports, fingerprints, now facial recognition. So…
We live in a digital world where companies like Facebook and Google have a lot of data about who you are and what you like….