Reclaiming Agency in the Age of AI
Why the smartest move in an AI-driven world might be building a model that knows you.
In the rapidly advancing world of artificial intelligence, there's a growing tension between embracing innovation and preserving human autonomy. As AI systems become increasingly integrated into our daily lives, concerns about losing control over our personal data, decision-making processes, and even our identities have intensified. The question isn't just about how AI will change our lives, but how we can actively shape that change to reflect our values and needs.
Enter the concept of Personal Language Models (PLMs)—AI systems trained on an individual's data, preferences, and communication style. Unlike generalized AI models that serve a broad user base, PLMs offer a tailored experience, acting as extensions of ourselves rather than external tools. They represent a shift from passive consumption of AI-generated content to active participation in AI development, allowing individuals to harness technology that truly understands and represents them.
This personalized approach to AI isn't just theoretical. Chris Do, an Emmy award-winning designer and founder of The Futur, collaborated with technologist Sho Rust to create the "Do Bot"—an AI trained on Chris's own content and communication style. This endeavor wasn't merely about automating responses; it was about capturing the essence of Chris's insights and making them accessible in a scalable, interactive format. As Chris noted, interacting with the Do Bot felt like "talking to myself," highlighting the profound potential of PLMs to mirror and extend our cognitive processes.
The implications of PLMs extend beyond individual convenience. In professional settings, they can streamline workflows, preserve institutional knowledge, and provide consistent communication aligned with organizational values. For creatives and entrepreneurs, PLMs offer a means to amplify their voice and maintain authenticity across various platforms and audiences. Moreover, by retaining control over their AI's training data, users can ensure that their PLMs evolve in alignment with their personal growth and changing perspectives.
However, the rise of PLMs also prompts critical discussions about privacy, data ownership, and the ethical use of AI. As individuals take on the role of both user and trainer of their AI counterparts, the responsibility for ethical considerations becomes more personal and immediate. This dual role empowers users but also necessitates a deeper understanding of AI's capabilities and limitations.personal.ai+3personal.ai+3deloitte.wsj.com+3
As we navigate this frontier, the goal isn't to resist the tide of AI advancement but to steer it in a direction that upholds and enriches the human experience. One way to do that? Train a model that thinks like you—not like the internet.
That’s the idea behind Personal AI—founded by Eggs! The Podcast alum Suman Kanuganti, a builder at the intersection of emerging tech and real human needs. Suman isn’t just theorizing about the future—he’s coding toward it, one personal language model at a time.
Reclaiming Your Voice, One Model at a Time: Suman Kanuganti
Suman Kanuganti has spent his career building technologies that don’t just automate work—they expand what it means to be human. Long before the hype around large language models, Suman was already focused on the interface between intelligence and identity. In 2015, he co-founded Aira, a groundbreaking assistive tech platform that used AR and human guides to help people who are blind or low vision “see” the world around them. Aira earned nods from Time and Fast Company, but more importantly, it showed Suman something enduring: that technology works best when it’s deeply personal.
That same principle drives his latest venture, Personal AI—a startup developing what he calls “personal language models,” or PLMs. Instead of scraping the internet, these models are trained exclusively on your own data—your tone, your thinking, your experiences—and operate under your control. With a background in robotics, blockchain, and early streaming tech, plus an MBA to bridge product and business, Suman brings a rare mix of deep technical fluency and human-first perspective. His goal isn’t just to build better tools. It’s to make sure that as AI evolves, the individual still gets a say in who they are—and how they’re remembered.
Building AI with a soul — and what you can do with it
The best founders don’t just invent new tools. They spot the invisible trends shaping how we think, work, and connect—and then build with those forces, not against them. That’s exactly what Suman Kanuganti has done with Personal AI. Across our conversation, he laid out not just a technical roadmap, but a human one: a way to coexist with artificial intelligence without losing ourselves in the process. Below are six key takeaways—each paired with a quote and a next step for anyone who wants to create smarter systems, protect their voice, and stay human in the loop.
"Technology works best when you start with the problem, not the tool."
Actionable Insight: Don’t start with “What can AI do?” Start with “What’s still broken?” Whether it’s how your team stores knowledge or how customers engage with your product, let the pain point—not the hype—guide your roadmap.
"Your memory used to live in your brain. Now it’s scattered across platforms."
Actionable Insight: Build systems to preserve—not just produce—knowledge. Consider how your personal or team-specific data could be turned into searchable memory: onboarding guides, client history, past decisions. AI isn’t just for speed—it’s for recall.
"Ethics starts with ownership. If your AI is trained on your own data—and you own that data—then the ethics of the model are a reflection of you.”
Actionable Insight: As you adopt AI tools, scrutinize the data sources. Ask: Whose data is this? Who owns the output? Favor systems that reflect your values and preserve your voice, rather than aggregate and anonymize it.
"If everyone gets their own AI, history changes."
Actionable Insight: Personal AI isn’t just about productivity—it’s about legacy. Whether you’re a founder, writer, or parent, start capturing your own thoughts and decisions now. The archive of you is already forming—make it deliberate.
"General AI is built to replace you. A personal AI is built to support you. The idea isn’t to replace humans—it’s to augment them or make people stronger.”
Actionable Insight: Use AI to make space, not noise. Let it handle routine responses, surface forgotten details, or simulate your voice—so you can spend more time on work that truly requires your presence.
"The interface doesn’t need to be smarter. You do."
Actionable Insight: Shift your thinking: AI shouldn’t feel like a foreign interface. The best systems amplify your conversations, not replace them. Look for tools that make you more coherent, not more generic.
Don’t Just Keep Up—Take Control
There’s no slowing the march of AI. The question isn’t whether it will change how we work, think, or communicate—it already is. The real question is whether we’ll let that change happen to us or choose to shape it ourselves. For most of us, the answer won’t come through building the next ChatGPT. It will come through smaller, more intimate choices: what we record, what we automate, what we claim as our voice—and what we don’t.
That’s the promise of a personal language model. Not just faster content or more efficient delegation, but a deeper kind of authorship: the ability to embed your perspective into the tools you use every day. You can offload your to-do list. You can even offload your memory. But you should never offload your identity. In a world of general intelligence, the personal is your edge. Train it like it matters—because it does.
Thanks for reading,
—Ryan
Ready for more?
Catch Suman Kanuganti’s interview in its entirety on Eggs! The Podcast.
Don’t miss a show! Subscribe on Spotify, Apple Podcasts, or really anywhere great podcasts are found.
Path Picks
Cool stuff to help you forge your path to greatness.
Note: The Path Weekly is reader-supported. As such, I may be using affiliate links below. If you want to support the newsletter at no additional cost to you, please consider using the links below. If you’d rather not, most items below are widely available anywhere you want to shop. Thanks! –R
Reading list
If you're looking to go deeper on the themes from this week's newsletter, here are a few books that pair well with the conversation and offer a broader perspective on learning, systems, and the future of work:
The Extended Mind by Annie Murphy Paul:
Why our thinking doesn’t stop at the skull—and how tools, environments, and even conversations extend cognition. A fascinating case for offloading mental work in smarter ways.Tools for Thought by Howard Rheingold:
Originally written in the early '80s, this book predicted many of today’s AI and computing challenges. A foundational text on personal computing as cognitive augmentation—well worth revisiting.Data and Goliath by Bruce Schneier:
A practical, readable overview of who owns your data, what they do with it, and why privacy matters in the age of surveillance capitalism. Required reading if you’re serious about owning your digital footprint.
More to explore
Suman Kanuganti Websites:
https://www.personal.ai/
https://www.linkedin.com/in/kanugantisuman/
https://x.com/sumanpersonalai
Work with me
Ryan Roghaar - Fractional CMO/Creative Director/Art Director: https://rogha.ar/portfolio
R2 - Creative Services for Agencies and SMBs: https://www.r2mg.com
Eggs! The Podcast: https://www.eggscast.com
Would you like a personal introduction to any of the incredible leaders featured in The Path Weekly to explore business or other collaborative opportunities?
Contact me here to learn more about my B2B matchmaking service.
Get featured
Do you want to be featured in a future edition of The Path Weekly?
Contact me to learn more.