The Experts (and ChatGPT) on Technology Trends, Algorithms, and AI: Part 2
By Sarah Birnie, Vice President of Analytics at Avalon Consulting
In Part 1 of this conversation in the DMFA print newsletter, we learned how leaders have been using AI — and now we’ll hear new, innovative ways teams are putting machine learning to good use.
The experts:
Roger Hiyama | Executive Vice President, Solutions & Innovation | Wiland
Susan Paine | Vice President, Data Analytics & Strategy | Human Rights Campaign
Derek Drockelman | Vice President, Sales & Marketing | ROI Solutions
Emily Courville | Senior Director, Analytics | The Humane Society of the United States
ChatGPT
What new ways are you now using AI?
Roger: Our Post-Merge Optimization (PMO) models identify the “bottom or worst performers” within a mail plan — in most cases, we can identify anywhere from 6% to 10% of bad names on nearly every mail file. Our PMO Savings Calculator shows with great transparency and high predictability which names should be dropped prior to mailing. The savings can then be re-deployed in more strategic ways — it’s like finding a new source of funding!
Susan: We partnered with Civis to leverage their AI-powered “ID Resolution” tool to ensure that we have one record for each individual across all platforms. This is the most sophisticated use of AI to date for our team — and someday, I expect it will start thinking on its own and become the “AI” that everyone is obsessed with these days!
Derek: We have been adding multiple dimensions to models. Our Ask Amount model predicts the response for each constituent and the likelihood of giving a bigger, smaller, or same-size gift. We are also integrating and enhancing our models with the expanded datasets available for clients using the Unite Analytics product, which acts as a central data repository across the organization.
Emily: We’re testing CoPilot and Scribe to distill notes and action items from meetings. We’re also testing ChatGPT to ask questions about coding in SQL and DAX. On the ROI side, we’re using their Machine Learning (MiLo) generated models to help optimize segment selects for mail by seasonality and package in appeals, and to determine what reinstatement names to select.
ChatGPT: Some of the new ways AI is being used include:
- Natural Language Processing (NLP): AI models like GPT-3 are capable of generating human-like text and engaging in natural language conversations.
- Generative Adversarial Networks (GANs): GANs are a type of AI model used for generating synthetic data, images, and videos. GANs have applications in areas like image synthesis, data augmentation, and creating deepfake media.
- AI Ethics and Bias Mitigation: Efforts are underway to develop frameworks for responsible AI deployment, fairness, transparency, and accountability.
- Climate Change Mitigation: AI is being applied to address climate change mitigation, natural disaster prediction, and renewable energy optimization. AI-powered models can analyze climate data, optimize energy usage, and improve resource management.
Don’t miss the final installment — read about “AI fails” in Part 3!
