Using AI to Help Write an Innovation Standard (ISO 56007 Development, Part 2)

Using AI to drive the idea innovation management standard, ISO 56007, toward progress, via a lightbulb.

Can you use AI to help write an innovation standard? If you are trying to be innovative, can you leverage advanced artificial intelligence to help you craft a written document with experts from around the world in the middle of a global pandemic? Where would you insert such technology into what some might consider a fairly technical matter with huge amounts of human interactions?

The 56007 working group found several uses of applied AI. This article will describe 3 ways teams can apply cutting edge AI to increase productivity, enhance creativity, and develop cohesion. Despites its broad uses, these approaches were a focused team initiative.

Transcribing ANSI/ISO 56007 Committee Meetings With AI

As the vice-chair of the ANSI/ISO 56007 committee, Chris Kaufman made sure the group had clear agendas and action tasks that were being tracked. The first, of course, is the use of natural language processing (NLP). An advanced AI tool, Otter.ai, was attached to the committee’s Zoom calls, which transcribed every meeting. This enabled the team to review comments, annotate meetings, and surface issues in case they needed to respond to challenges to prior discussion. This behaves similar to a digital assistant that can easily mediate if meeting notes cause any questions to arise.

Drafting ISO 56007 With AI

ANSI/ISO 56007 committee chairman Robin Rowe wanted the team to try something innovative to write the innovation standard, to practice extreme innovation. “We’re using OpenAI GPT-2 to draft the first world standard document created by AI,” says Rowe.

The blank page is a writer’s greatest fear. Writer’s block delays progress. Once in draft, progress is faster. It’s easier being a critic. OpenAI GPT-2 is a complete-my-thought type of AI. Given the first paragraph, GPT-2 will continue it by writing what it anticipates what you would say next. https://github.com/robinrowe/gpt-2

Using OpenAI to draft a standard is a 3-step process. A group of domain experts define the topics to be covered using mind mapping software, such as open source FreeMind. A mind map is a connected set of topics, a spider graph. With that mind map as a guide, the committee can quickly write a concise one-paragraph intro for each named topic, then feed that into OpenAI GPT-2. It outputs one page per topic to produce a first draft. Subject matter experts then edit the draft as usual, looking for logic errors, omissions, and grammatical issues.

Innovative Technology Helping the Innovation Standards Committee

In addition to GPT-2, the innovation standards committee also deployed another innovative emerging technology tool.

The innovation standards committee had a repository of over 200 research papers to use in helping them write out the new standards. However, it was time cumbersome to go through all research papers and pull out relevant insights.

For this reason, the team decided to try a new NLP technology called EIVA (Enterprise Intellectual Virtual Assistant) created by LAUNCH!, a Los Angeles AI consulting and solutions startup. The solution helped the committee with on-demand accelerated and targeted learning in order to write a better paper in a shorter amount of time. EIVA was loaded onto the committee’s Slack channel, and committee members had the opportunity to ask EIVA any question they could think of regarding innovation. EIVA then used a deep learning based semantic search model, which is more advanced than basic key word search models, to surface the most relevant answers after querying the entire repository of unstructured data.

For example, when asking about what metrics are used to measure innovation, EIVA answered:

“Innovation metrics are important for at least two reasons.  First, metrics help managers make informed decisions based on objective data, which is especially valuable given the long-term nature and risk associated with certain innovation projects. Second, metrics affect behavior by helping align goals and actions with the best interests of the company. Among those companies that do measure their innovativeness, most use R&D and product-development metrics only, such as annual R&D budget as a percentage of annual sales, number…”

This gave the team a starting point on where to focus their section.

EIVA helped the team accelerate the research behind creating the innovation standards and ensure they were tapping into the available knowledge.

Thus, despite a global pandemic, stressed resources, and difficult deadlines, the 56007 group was able to deliver a standard draft in less than a year. AI is not a panacea for hard work, constructive collaboration, or deep thinking. But the tools do exist today to harness specific aspects of the technology to fill the gaps in our ever challenging and demanding world.

This is the second post in a series written by the 56007 working group focused on the development of the innovation idea management standard. You can read the first entry, Can Innovation Be Standardized?, here.

Contributing Author: Robin S. Rowe 

Innovation director, research scientist, C++ software architect and tech journalist (Popular Science, Popular Photography). Innovation provider to Lenovo (AR glasses), AT&T (digital video), GoPro (IoT/cloud), DreamWorks Animation (renderfarm), DARPA (AI), the navy (VR war gaming) and NBC-TV (robotic studios). System architect of animation technology Disney Marvel uses to make major motion pictures and Mattel uses to produce a hit Barbie cartoon series. 

Chairman ANSI/ISO 56007 innovation idea management. ANSI ICAC voting  member. ISO 21030 continuing education conformity assessment JWG expert. ISO 17021-13 management systems conformity assessment JWG expert. Chairman CFO Alliance cyber security subcommittee. Winner Novartis Biome innovation prize. Multiple HeroX innovation awards. 

Contributing Author: Christopher David Kaufman

Christopher David Kaufman is an author, researcher, and practitioner in technology innovation and sales transformation. 

He is currently working on his doctorate at the Vanderbilt in organizational management and learning systems. He is the co-chair of the Idea innovation management working group for the ISO 56007 standard. His work focuses on how to utilize the latest research and technology to transform sales operations into innovation machines that propel the organization. He has worked as a software developer, entrepreneur, VP of sales, executive sales trainer, sales consultant, and chief strategist for a Fortune 100 firms. He has authored books and created patented technologies. To learn more how you can help define how the world innovates ideas, you can reach him at cdk64@cornell.edu or on linkedin.com/in/cdkaufman.

Contributing Author: Rebecca Bennett

Rebecca Bennett is an entrepreneur, technology consultant, author, and podcast host.

Rebecca is the founder of LAUNCH! an AI consulting and solutions startup in Los Angeles. She is also a member of the Idea innovation management working group for the ISO 56007 standard. Her work as a technology consultant helps organizations realize the value proposition of intelligent and automated technologies, while finding innovative ways to apply it. Formerly she was driving the digital transformation for a major FinTech organization while earning her MBA at the USC Marshall School of Business. She also hosts a business podcast with a focus on leadership which you can listen to at https://launchitspot.com. You can learn more about Rebecca at https://www.linkedin.com/in/rebeccabennettpmp/

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.