Hmmmmm. After nearly eight months of indulging in AI “jamming,” my newly coined term for recreational learning plus a few personal projects with my nephews and grandkids, I've begun to uncover potential obstacles to scaling AI change management methods in organizations. It stems from over 5 years of casual but deep learning (and teaching!) data analytics and data science in various webinars and tutoring.
“Creating better humans will always be more important than creating smarter machines!".
Garry Kasparov, former world chess champion and author of "Deep Thinking", opened the OECD conference Artificial Intelligence: Intelligent Machines, Smart Policies" on 26 October 2017.
Last Sunday, I found myself engaged in a discussion with a programmer and AI enthusiast who could potentially become my video channel partner. We were grappling with naming our channel. For more than an hour, we tossed ideas back and forth. Engaged in brainstorming sessions. And debated the merits and demerits of each suggestion. In the midst of it all, we marveled at how AI continues to captivate our imagination with its almost limitless possibilities. We also chuckled at some of the ideas presented by those who oppose AI. We were amazed with the spread of AI-powered software. AI Chrome extensions, plug-ins, “functions,” and even AI co-pilots. In fact, much unlike the software Apps of the bygone days that took years to launch, there’s a new attention-grabbing AI born daily. Today! However, we kept coming back to the name for our video channel. Ultimately, we settled on a tentative name! But agreeing to revise it later.
Monday came, and I awoke with a striking thought: Why did two AI aware individuals like us and chatting about AI, not use AI to solve a simple brainstorming exercise? You'd think it would be by force of habit. I can hear Kurt Lewin's whisper: Unfreeze, Change, Refreeze. Forgetfulness seemed an unlikely explanation too. Considering that my friend is younger than 25.
Later on, I turned to ChatGPT 4.0 and posed the problem to it. In less than five seconds, it churned out 100 alternative names. Hold that thought for a moment. Can you imagine the immense boost in productivity if this were to occur within an organization? It's mind-boggling, to say the least. Bordering on the miraculous!
Micro Processes
Upon reflection, I concluded that despite the AI disruptions around us, our mindsets remain rooted in the 90s. We fail to adopt the new AI mindset yet. Even if our thought leaders continue to debate whether AI is a friend or foe. This was the same feeling I had when I initially started using Google Search for seeking second opinions. It took me approximately 2 years, give or take, for the change to truly sink-in and transition from my old ways. Consider the key milestones in the change process. Exploration (testing the efficacy of online search). Running parallel processes. Establishing trust in the new method. Forming habits around its usage. Making the switch. Acquiring and mastering advanced techniques. Further building trust. And being ready for further changes. Now, envision scaling these micro processes within your organization. The challenge for OD practitioners becomes evident. Shifting paradigms requires an excruciatingly long time. Double (or even triple) your own original estimates. Unless these new micro processes are eliminated, condensed, or simplified, change processes will be hindered.
Practice is Faster than Theory!
Rommel Ancheta, an experienced practitioner in talent and OD, astutely observed that practice outpaces theory when it comes to AI. Our practice today will be our theories of tomorrow, he said. So, we need to accelerate how we relate, work with and apply AI to scale it’s use in our organizations.
This sentiment resonates with Cassie Kozyrkov, Google Cloud's Chief Decision Scientist, who introduced Decision Intelligence to Google and personally trained over 15,000 Googlers. She offers valuable advice on maintaining an "AI safety mindset," for organizations. She reminded practitioners that the objective will always be subjective. The goals you set for AI to learn will depend entirely on how you intend to use it in your organization.
To truly accelerate AI deployment, OD practitioners ought to immediately sidestep the gobbledygook’s. Black Box. Machine Learning. Deep Learning. Algorithms. Instead, use it as you would a new appliance, say a bread toaster or a washing machine. Forget the manual. Seek out the minimum instructions. Give it a spin. Persist until it becomes second nature!
AI4OD
In this spirit, SAIDI is organizing 2 consecutive six-hour Saturday webinars on AI4OD: Transforming the Future Organizations Today. Pitched for internal OD practitioners in government and corporates, the course will foster a deep dive on AI applications and broadly sketch some guides in scaling its deployment.
The content will cover concepts, applications, and ethical considerations relevant for the practitioner. There will be few lectures. Extensive demonstrations and practical exercises involving real-world instances in OD, HR, planning, operations, marketing, and data intelligence. As the course ends, the participants will integrate AI into OD interventions and change strategies. They will craft concrete action plans to leverage AI in propelling organizational change and fostering innovation.
Making AI Safe
Without being a spoiler, let’s bring Cassie Kozyrkov back. I really recommend watching her presentation till the end. She recommends several ways of making AI safer for us and our organizations.
1. Forget science fiction. It's a dangerous distraction.
2. AI is built tested and designed by humans.
3. AI is a tool designed to transform inputs into outputs. A productivity tool!
4. How ever captivating it may be, its still a machine.
5. Traditional programming requires instructions. AI uses examples.
6. AI is useful in situations where it's easier to come up with examples than instructions.
7. The purpose of an AI system matters in determining its function.
8. There is no one objective answer in AI, it depends on the project's purpose.
9. A system built for one purpose may not work for a different purpose.
10. And if I may add….do no harm!!!
And now, let’s watch Cassie together!
NOTES:
1. SPECIAL INVITATION TO OUR READERS IN THE PHILIPPINES. You are invited to share your opinion on the HR/OD Trends for presentation and discussion in our forthcoming ODLab 2023 in Davao City on July 20 to 21. Share your voice here.
2. Event Announcement: ODLAB 2023: REIMAGINE OD. ODPN takes the lead in driving and inspiring a new direction for OD in the Philippines. Join us on July 20—21, 2023 in Davao. Connect to ODPN (odpnsecretariat@gmail.com) for a brochure, registration and sponsorship opportunities. Join us! the date!!!
3. Our next issue will be about our ODLab2023. Join us. Contribute your thoughts for the the next issue.
4. Forthcoming ODPN Courses for OD Professionals
Dr. Joy Calleja OD Diagnosis, September 6 (In-Person)
Dr. Josephine Perez Process Observation and Analysis, October TBA
Dr Melissa S Reyes Designing OD Interventions November 16 (In-Person)
5. Dr. Ed Canela’s Forthcoming Course
AI for OD Practitioners, July 15 and 22, 2023, (2 Saturdays, online, 9:00 am to 5:00 pm) for Php 5,000 Southeast Asia Interdisciplinary Development Institute (SAIDI) Graduate School of Organizational Development and Planning. Email interest to Ms. Cynthia Cruz.