Humanistic AI: Merging Technology with Human-Centered Ethics and Skills
What skills can help us get the most out of AI?
When we approach AI in a human way, with people at the center, it can be a powerful tool that augments how we work. As AI becomes more widespread across industries, it’s critical to learn the skills to harness and collaborate with AI in a way that is human-centered, ethical, and impactful.
In this episode of the Creative Confidence Podcast, we speak with Tom Gruber—product designer, entrepreneur, and co-creator of Siri. He shares critical skills to get the most out of AI, why AI needs human-centered design, and the ethics of using AI tools.
Join an August cohort course for exclusive access to our live AI x Design Thinking workshop series.
JUMP TO SECTION
Why Generative AI Needs Human-Centered Design
Collaborating with AI: Why Humanistic AI Matters
Critical Skills for Working with AI
Learn more about how to bring a human-centered approach to AI with our limited-time AI workshop series—enroll in any June cohort course or certificate by June 20 for exclusive access.
Why Generative AI Needs Human-Centered Design
Tom emphasizes the importance of combining technical knowledge with the skills and mindsets of human-centered design when working with emerging technologies such as generative AI. Having co-founded multiple startups, he learned that many products fail because they are too technology-driven without considering human needs. He now advocates for a human-centered design approach that keeps the user’s experience at the center.
AI inherently optimizes for certain aspects or qualities—for example, past technologies have optimized for user engagement on social media. Tom stresses the importance of design thinking and systems thinking in AI, focusing on the human needs being met. He says that if we only consider the engineering side of things, we risk unintended consequences, such as addiction in the case of social media. Bringing in human-centered design can help us use and design AI in a way that supports and enhances human capabilities more effectively.
What is Generative AI?
At its core, generative AI can understand and generate language, reading and comprehending entire documents and conversations. Its capabilities extend to other creative fields and modalities as well, such as generating images and videos from written descriptions.
Generative AI originated from efforts in language translation with large language models, or LLMs. These models were trained by processing vast amounts of text from the internet, learning to generate sentences based on preceding words. This method allows AI to capture a wide range of our human cultural legacy, effectively compressing the internet's information into a usable model for generating coherent text.
Tom says that when generative AI creates text, it functions like an advanced autocomplete, predicting and producing words based on input patterns. These models have been trained to such a high level that they can generate text by understanding and generalizing from vast amounts of data.
“Human-centric AI is about, where does the AI live and where does the human live? What's the partnership? Design thinking and systems thinking are methods for thinking about how technology and humans work together.”
Tom Gruber, Product Designer, Entrepreneur, & Co-Creator of Siri
Collaborating with AI: Why Humanistic AI Matters
People have differing views on AI—some are cautious and fearful, while others are enthusiastic and optimistic. Tom believes the key lies in balancing these perspectives and recognizing AI’s potential as a valuable knowledge partner and source of collaboration that keeps humans at the center.
Tom explains the concept of humanistic AI, which is about creating AI that benefits people, individually and collectively. Unlike a technology-centric approach that focuses on automating humans, humanistic AI aims to help humans do what they do best. While some roles may ultimately be replaced, the goal for AI is to enhance human life through health, education, and other areas. Tom compares it to glasses augmenting our vision without replacing our eyes: AI should augment our intelligence without replacing our thinking.
Tom describes Generative AI as a “language savant”—it functions as a sophisticated language machine, generating text and images based on learned patterns without truly understanding or verifying the information. AI lacks an understanding of epistemology, not knowing what is true or false. Because of that, AI has what Tom calls "hallucinations," where AI provides confident but incorrect answers.
Learn more about human-centered design methodologies in our online Foundations in Design Thinking Certificate.
Critical Skills for Working with AI
The key AI skills for the future aren’t just in prompt engineering. Because of AI’s limitations and lack of epistemology, Tom stresses the importance of critical thinking. He thinks of AI as a team of agents rather than as oracles of truth—AI will often contradict itself, making it important to know how to manage AI. Tom compares managing AI to mentoring students or leading a creative team, where diverse opinions must be synthesized. He also suggests assigning specific roles to AI, such as a skeptic or a creative, which mirrors how a team functions.
Tom points out how humans adapted their search queries with Google over time, becoming more precise and detailed. In the same way, coevolution will continue with AI, as people and AI systems learn from each other, improving the quality and relevance of AI responses. Many organizations are already integrating AI into their systems, and the data from queries will help AI models better understand and respond to human needs. This dynamic relationship will require people to stay informed and adaptable to keep pace with the fast-moving developments in AI technology.
Additionally, Coe adds that sense-making, a form of critical thinking or systems thinking, is increasingly important. It’s not just about knowing how to use AI tools, but understanding when and how to use them to solve human problems strategically. This encourages thinking about the bigger picture and integrating AI into broader problem-solving frameworks. With a human-centered approach to AI, we can harness the power of AI as a collaborative partner and powerful tool to augment our skills and tackle complex challenges.
“We have the possibility to have a human-centric augmentation and collaborator that we've never had before. Truly making us our better selves—not by replacing us in the chair, but by sitting with us.”
Tom Gruber, Product Designer, Entrepreneur, & Co-Creator of Siri
Ethical Guidance for AI
For people who develop or work with AI tools, it’s important to consider how to use them ethically and responsibly. At IDEO’s Emerging Technology team, we’re thinking about the intersection of AI & ethics. Coe outlines a few key ethical guidelines when using AI tools:
- Know the Tool: Understand the training model including its biases and limitations. Be aware of what creators might be cut out or plagiarized by the tool. Pay attention to privacy settings and how inputs might train the model.
- Consider Context: Evaluate the ethics of the context in which the tool is used. For instance, using AI for vacation tips is fine, but synthesizing performance management data with an employee's name is not appropriate. Always consider the potential harm.
- Attribute Properly: Always provide attribution when using AI-generated content. Recognize AI as an augmentation of human capacity, not a replacement.
Tom also stresses the importance of setting appropriate goals when training AI systems. For example, if the goal is to predict job success or the likelihood of someone getting promoted, it's crucial to carefully choose optimization criteria. If not done correctly, the AI might produce biased results, such as favoring certain demographic groups, which doesn't equate to better candidates.
Additionally, many AI models generate patterns based on their training data without a true understanding of the information. He warns against giving AI systems control over complex operations, such as financial services or programming environments, just because they can generate coherent language. These systems were not trained for these tasks, and it can lead to serious ethical and practical issues.
Tom uses the metaphor of a talking chimpanzee to illustrate this point: even if it seems intelligent, it doesn't mean it's capable of handling complex tasks like flying an airplane. There is a risk of over-reliance on AI for tasks it wasn't designed to perform. By focusing on careful goal-setting and understanding AI's limitations, teams can use AI tools responsibly and ethically, so they don’t cause unintended harm.
“Helping humans do what humans do is a humanistic goal, so the thing you’re trying to achieve is advancing and augmenting human life instead of automating them out of their own life.”
Tom Gruber, Product Designer, Entrepreneur, & Co-Creator of Siri
The Future of AI
According to Tom, AI has the potential to be as transformative to humans as fire or electricity, with the ability to revolutionize the ways we work, learn, and interact with the world. Tom highlights the application of generative AI in the field of education. He imagines having a personal tutor on our side of the screen, helping us with tasks, providing advice, and nurturing our development. AI could democratize access to personalized education and continuous learning, making it possible for everyone to have a tutor tailored to their individual needs.
However, Tom also notes the potential risks of AI. One of the most pressing issues is the potential for fraud. AI can be misused to make deep fakes, spread misinformation, and create propaganda, which pose serious threats to trust and authenticity on a macro level. Tom says that generative AI can be a force for good, but its potential for misuse also means a careful and ethical approach is critical.
This summary was synthesized with the assistance of generative AI.
About the Speaker
Tom Gruber
Product Designer, Entrepreneur, & Co-Creator of Siri
Tom Gruber is a product designer and entrepreneur who uses AI technology to augment human intelligence, individually and collectively. He was cofounder, CTO, and head of design for the team that created Siri, which was purchased by Apple in 2010.
Tom’s professional career spans four decades, from graduate school during the classical AI period, through research at Stanford during the dawn of the Web, through founding a series of companies, to his present incarnation as speaker, writer, and impact advisor. Throughout his career Tom has pursued the question of how technology can help us live better lives. Applying ideas from artificial intelligence, cognitive science, and design, Tom’s work explores how connecting people and machines can foster collaboration, learning, knowledge sharing, and getting things done. Learn more about Tom's perspective on Humanistic AI.
If you want to learn more about how to bring a human-centered approach to AI, check out our limited-time AI workshop series—enroll in any June cohort course or certificate by June 20 for exclusive access.
- choosing a selection results in a full page refresh
- press the space key then arrow keys to make a selection
