AI innovation labs

Adapting proven strategies for a new era

Since I published my book “Innovating” last year, business leaders have often asked me about innovation labs. In particular, they have wondered if my advice in the book would differ for new AI labs they would like to set up.

A particular concern is that many organizations need more technical skills to explore issues such as model pre-training, fine-tuning, or customization. However, they still want to explore AI’s implications for their business. So, how might they do this?

Firstly, AI innovation labs do present a practical way for organizations to approach AI from a strategic and exploratory perspective rather than solely focusing on the technical aspects of AI development. While development skills are undoubtedly valuable, they are not an absolute prerequisite for successfully leveraging a new technology in business.

The rise of AI-driven design tools, no-code and low-code AI platforms, chatbot builders, augmented analytics tools, and, for developers, cognitive services APIs has lowered the barrier to entry for AI solutions. Everyday AI tasks, such as image creation, text generation, image classification, sentiment analysis, conversational AI assistants, and predictive modeling, are now within the reach of modestly skilled teams.

This is what AI thinks an AI Innovation Lab would look like. I expect yours will be a little different.

However, while these tools lower technical barriers, it’s still crucial for organizations to develop a strategic understanding of AI and its limitations. This is where AI innovation labs can play a vital role. Moreover, as organizations progress, they may need to gradually build their internal AI development capabilities or partner with external AI experts. AI innovation labs can help by identifying specific skill gaps, training and development opportunities, and providing a hub for collaboration with AI specialists.

New strategies

In Chapter 21 of Innovating, I set out a simple list of priorities for an innovation lab:

• Define the purpose and objectives of the lab

• Assemble a diverse team with various backgrounds and skill sets

• Establish a governance structure that balances guidance and autonomy

• Allocate resources and create a physical or virtual space conducive to collaboration

• Define a straightforward process for the creativity, evaluation, and implementation of ideas

By focusing on exploratory and strategic aspects of AI, innovation labs can help organizations develop a sound understanding of how AI can improve processes and create new opportunities. They can bridge AI's technical complexities and the organization's practical realities, translating AI concepts into actionable insights and strategies.

Here’s how these priorities may change for an AI-focussed lab:

Clarify AI-Specific Objectives and Ethical Guidelines:

• Define the lab's goals in terms of understanding, exploring, and applying AI within the organization's context

• Establish a solid ethical framework to guide the lab's activities, considering data privacy, algorithmic bias, and transparency.

Build an AI-Literate Multidisciplinary Team:

• Assemble a team with diverse backgrounds, including domain experts, strategists, and AI enthusiasts, even if they lack deep technical skills

• Provide access to training and resources - online training can be excellent and valuable - to help team members develop a foundational understanding of AI concepts and technologies.

• Encourage collaboration and knowledge sharing within the lab (of course) and with other teams, even in the exploratory phase. The AI world is changing rapidly, and new developments continuously catch the attention of every department. It’s difficult for one team to keep up!

Collaborate with AI Experts and External Partners:

• Engage with AI specialists, researchers, and consultants to gain insights and guidance on AI technologies and best practices

• Participate in AI industry events, workshops, and communities to keep up with relevant developments and connect with potential partners.

• Explore opportunities for collaboration with AI technology providers, research institutions, and other organizations working on similar challenges.

Embrace Experimentation and Rapid Prototyping:

• Encourage a culture of experimentation and iteration, focusing on quickly testing ideas and learning from failures

• Utilize no-code or low-code AI tools and platforms to enable rapid prototyping and concept validation

• Regularly showcase and share the lab's experiments and findings with the broader organization to gather feedback and spark new ideas.

Prioritize Explainable and Human-Centered AI:

• Focus on developing AI applications that are transparent, interpretable, and aligned with human values and needs

• Engage end-users and stakeholders throughout the ideation and development process to ensure that AI solutions are practical, intuitive, and valuable.

• Continuously assess and address potential ethical risks and biases associated with AI applications.

Plan for Scalability and Organizational Integration:

• Design AI experiments and prototypes with scalability and integration in mind, considering how they can be expanded and implemented across the organization

• Collaborate closely with IT, operations, and other relevant departments to identify opportunities for AI integration and ensure smooth deployment.

• Establish clear metrics and success criteria to measure the impact and value of AI initiatives.

What could an AI lab do?

As I suggested, AI labs can bring together a diverse group of individuals, including domain experts, business strategists, product managers, and AI enthusiasts (who may be found scattered throughout an organization regardless of job title), to collectively explore future uses of AI.

These multidisciplinary teams can engage in various activities to explore AI's potential, such as:

• Conducting industry-specific research and analysis to understand the current state of AI technologies, trends, and relevant best practices

• Identifying potential use cases and opportunities for AI within the organization, considering factors such as business value, feasibility, and alignment with company goals

• Engaging with external AI experts, consultants, and technology providers to gain insights, guidance, and support in navigating the AI landscape

• Experimenting with AI tools, platforms, and methodologies to develop a hands-on understanding of their capabilities and limitations

• Developing proof-of-concepts and prototypes to validate ideas and demonstrate the potential value of AI applications within the organization

• Collaborating with internal stakeholders to assess the organizational readiness for AI adoption, including considerations of data availability, infrastructure, skills, and culture

By focusing on these exploratory and strategic aspects of AI, innovation labs can help organizations develop a sound understanding of how AI can improve processes and create new opportunities. They can bridge AI’s technical complexities and the organization’s practical realities, translating AI concepts into actionable insights and strategies.

Moreover, AI labs can catalyze the development of AI literacy and foster an organization’s innovation culture. Labs can help demystify AI and encourage a broader understanding of its potential impact by engaging employees from various departments and backgrounds in AI exploration. This, in turn, can lead to more informed decision-making, increased adoption of AI technologies, and a more agile and adaptive organization overall.