Former Google AI Founder and Machine Learning PhD Reveals How Leadership Has Transformed — And What Leaders Must Prioritize Now
AI-first leadership means letting go of control: Modern leaders must shift from micromanaging decisions to orchestrating direction — trusting AI systems to handle data, tasks, and optimization so humans can focus on creativity, empathy, and strategy.
Work is becoming skills-based and fluid, not role-based: AI breaks work into tasks, enabling dynamic teams that form around problems instead of static job titles or hierarchies. Leaders now enable talent flow rather than manage rigid org charts.
AI-ready organizations experiment fast and measure impact: Success comes from embedding AI into real workflows, running rapid pilots, and tracking measurable outcomes — not from demos or standalone tools. Culture, speed, and experimentation drive AI ROI.
In our interview with him, he shared how he's creating an AI-first organization, and how it's changing both his role and his leadership style.
Ashutosh Garg’s journey from AI researcher to AI-driven leader
My academic background laid a strong foundation for my interest in AI. I pursued a Ph.D. in machine learning from the University of Illinois at Urbana-Champaign, where my dissertation focused on learning in high-dimensional spaces. This work earned me the Robert T. Chien Award for excellence in research, and it was during this time that I developed a deep fascination with the potential of AI to solve complex problems through algorithms and data analysis.
My early career experiences at IBM and Google further solidified my interest in AI. At Google, I was part of the research team responsible for personalization efforts and core ranking algorithms, which are crucial components of AI-driven services. This role allowed me to see firsthand the transformative power of AI in enhancing user experiences and optimizing search functionalities.
Keep Reading—and Keep Leading Smarter
Create a free account to finish this piece and join a community of forward-thinking leaders unlocking tools, playbooks, and insights for thriving in the age of AI.
Already a member? Log In
The entrepreneurial bug bit me when I cofounded BloomReach, a company that leveraged big data and AI to enhance digital experiences. This venture taught me the importance of applying AI to real-world problems, particularly in the e-commerce space.
And later, I founded Eightfold AI, driven by the mission of using AI to help people find meaningful work. The idea was to create an AI-powered talent intelligence platform that could revolutionize how talent is managed across the globe. We enable the right career paths through AI by addressing biases, improving skills matching, and enabling personalized career paths. The platform's ability to process vast amounts of data to provide insights into career development and job matching is a testament to the power of AI in transforming HR processes.
We're using AI to create a more inclusive workforce, which aligns with my belief that technology should serve to uplift and empower individuals.
Why AI-first leadership requires letting go of control
I’ve always prided myself on being someone who likes to “take the wheel.” I like to drive, to have control — whether that’s literally behind the wheel of a car or in leading a company.
For a long time, I saw control as essential to good leadership. But the first time I sat in a Waymo — an AI-enabled self-driving car — it changed my perspective. I realized very quickly that I was in one of the safest, smoothest, most precise driving experiences imaginable — and I wasn't the one in control. The system was making hundreds of micro-decisions every second, drawing from millions of data points, and doing it more consistently than any human could. That moment made me think differently about leadership in an AI-first world.
AI, like Waymo, doesn’t take control away — it redefines it. It frees leaders from constantly “driving” every decision so we can focus on the direction, not the steering. In many ways, leadership today is about learning to trust intelligent systems, letting them process the data, make the optimizations, and surface the insights that allow humans to focus on creativity, empathy, and strategy.
That’s a shift I’ve brought into how I lead at Eightfold. We’re building systems that learn continuously, that make people better at what they do, and that help leaders let go of micromanaging in favor of enabling.
Just like in a Waymo, once you get comfortable letting AI take the wheel, you realize it’s actually a premium experience — faster, safer, and far more freeing.
How AI is reshaping old models of organizational design
Everything about how we think of work and leadership is changing. In an AI-first world, the old models — hierarchies, static roles, rigid org charts — simply don’t make sense anymore.
AI is breaking work down into its smallest building blocks — tasks, not titles. Instead of thinking, “I’m a marketing manager” or “I’m a recruiter,” people are increasingly thinking, “Here are the problems I’m solving, here are the tasks I’m best suited for.”
That's a huge shift. It means organizations will become far more fluid. Teams will form and re-form around the work that needs to get done, rather than around fixed roles or reporting lines.
How AI is transforming leadership from managing to orchestrating
For leaders, that changes everything. My role is less about managing and more about orchestrating — making sure the right people with the right skills are focused on the right tasks at the right time. It’s about enabling, not controlling. And AI gives us the data and visibility to do that in real time. We can see what skills exist across the organization, where there are gaps, and how to redeploy talent dynamically.
We're using our own tools internally to do this, and early impacts include significant improvements in knowledge retention, onboarding speed, and collaboration. And for me, personally, it's been a game-changer in terms of getting real-time transparency when I need it.
Automation is also freeing people from the repetitive parts of their jobs. I find that so exciting because it lets humans focus on the work that’s actually satisfying, creative, and high-ROI. When people spend more time on things that energize them, not just things that fill their calendars, everyone wins — the individual, the organization, and the customer.
So yes, I’ve had to let go of a lot of old assumptions — that leadership is about structure, or control, or linear growth paths.
The new world of work is dynamic, skills-driven, and AI-augmented. Leadership is about creating the conditions for that system to thrive — empowering people, building trust, and helping them grow faster than the technology itself.
How AI is helping teams operate at a higher level across every function
What’s exciting about this new era of AI goes beyond automating tasks — it’s amplifying people’s abilities across every function.
Take our sales development teams, for example. AI now helps SDRs do in minutes what used to take hours. They can identify the right accounts, personalize outreach, and qualify leads with far more context. That means less time researching and more time actually engaging with prospects — and those conversations are deeper because they’re powered by real insights.
For engineers, it’s similar. With AI-assisted coding and testing, productivity has gone up dramatically. Engineers can prototype, debug, and ship code faster — but just as importantly, they’re spending more time solving meaningful problems rather than rewriting boilerplate or chasing down small errors.
Even something as simple as quarterly business reviews — which used to be a huge time sink — are completely different now. We’re using AI to pull together real-time performance data, analyze customer health, and surface trends automatically. Instead of spending days building slides, teams now spend their time discussing insights and actions.
That’s the real story of AI empowerment: It’s giving people back time and focus. It’s letting every team operate at a higher level — with more creativity, more precision, and more satisfaction in the work they’re doing.
Why AI success depends on designing organizations that experiment and measure
An MIT study recently showed that 95% of companies investing in AI aren’t seeing meaningful ROI yet — and that gap between the promise of AI and results doesn’t surprise me.
Too often, AI projects are launched as experiments, not embedded into core business workflows. They look impressive on a slide, but they don’t connect to measurable outcomes.
So, we focus on tangible, defensible value. As an example, we've seen a huge uplift in recruiter productivity and even candidate experience through our AI interview tool that we use internally within our recruiting teams. Every interview has a real, quantifiable cost in recruiter and hiring manager time, so we automated the process. The value isn't theoretical; it’s measurable.
The lesson is simple: AI only creates value when it’s deeply integrated into how work actually happens — not when it’s layered on top. As a leader, that means designing organizations that experiment fast but also measure impact relentlessly. We don’t celebrate demos; we celebrate results.
How to build an organization that experiments quickly and measures impact
To design this kind of organization, you have to first design a culture of experimentation. Teams need the freedom to try new ideas, and it has to be okay to fail — so long as you fail fast, learn, and iterate.
Second, the organization has to prioritize speed. In today’s environment, speed is not a luxury — it’s the competitive advantage. You need small, empowered teams, fast feedback loops, and the ability to ship, measure, and refine quickly.
Third, you must continuously design for innovation. Innovation doesn’t happen by accident. Every year, organizations should intentionally allocate a meaningful portion of their people and resources to new ideas, new products, and new bets.
If you don’t plan for innovation structurally, other priorities will crowd it out.
Trust people to embrace what works. We often assume AI will feel impersonal or cold, but when it’s built with empathy and clarity, it can actually create a better, more human experience.
What being an AI-ready organization really means
For me, being “AI-ready” means three things:
- Every team operates with an AI-first mindset.
- People are empowered to use and improve AI tools on their own.
- People treat AI as a teammate; not a threat.
Ultimately, the goal is simple: to make AI second nature. When people stop asking, “Should I use AI for this?” and start asking, “How can AI help me do this better?” — that’s when you know you’re truly AI-ready.
For us at Eightfold, AI literacy isn’t a training module — it’s a mindset. You can’t be an AI company unless every person, in every function, feels comfortable using and questioning AI every day. So, instead of treating AI as something separate, we’ve woven it into how we work.
We “dogfood” everything internally. Everyone at Eightfold uses our tools in their own jobs, whether it’s asking their digital twin for updates, using AI to prep for customer meetings, or automating follow-ups. It’s hands-on by design. The best way to build AI fluency is to live with it.
We also run regular AI learning sessions for non-technical teams. The goal is not to turn everyone into engineers, but to help them think about what’s possible. Once people see that AI isn’t magic — just math and data used in creative ways — the fear goes away and curiosity takes over. That’s what we want.
And we've launched internal programs to help people experiment with AI in real business workflows — in HR, marketing, finance, everywhere. We encourage small pilots that prove value quickly, because adoption happens when people experience the benefit themselves. What we call "Project Andromeda" is a good example.
AI only creates value when it’s deeply integrated into how work actually happens — not when it’s layered on top. As a leader, that means designing organizations that experiment fast but also measure impact relentlessly. We don’t celebrate demos; we celebrate results.
How Eightfold evolved from standalone tools to a unified agentic operating layer
Over the past year, we launched "Project Andromeda," a company-wide effort to deploy agentic AI across Eightfold. We identified 72 high-impact use cases and have already implemented many of them, making agentic workflows a core part of how we operate.
In research and development, tools like Cursor and GitHub Copilot have meaningfully increased engineering velocity. Our teams now spend more time on innovation and less on bug fixes, documentation, and on-call work, as agentic systems proactively handle much of the routine load.
In sales and marketing, our in-house Andro Bot supports discovery prep, scoring, and account planning, while agentic SDRs automate outreach and qualification. This has streamlined RevOps and improved consistency and efficiency across the funnel.
In general and administrative, we’ve implemented agentic workflows for HR, Legal, and Finance — from accounts payable to legal/compliance Q&A — reducing turnaround time and freeing teams from routine tasks.
Collectively, these changes have transformed our internal tech stack from a collection of tools into a coordinated, agentic operating layer. We’re using the same agentic principles internally that we deliver to our customers, which keeps us ahead of the curve and grounded in real operational impact.
Why leaders should trust people to embrace AI
Overall, I've been surprised by how willing people are to embrace AI.
It’s a good reminder for me as a leader: Trust people to embrace what works. We often assume AI will feel impersonal or cold, but when it’s built with empathy and clarity, it can actually create a better, more human experience. That realization has really reinforced how I think about AI design — not as a replacement for human connection, but as a way to make it more consistent, scalable, and fair.
As leaders, our job is to create environments that are adaptive, experimental, and always learning. It’s not about knowing every answer — it’s about building organizations that can keep up with the train once it starts moving.
The shift to AI is a bullet train — and why leaders can’t afford to miss it
And finally, I'll say this: The shift to AI is like moving from an analog train to a bullet train.
In the old world, if you were a minute late, you could still hop on — you might lose some time, but you’d catch up eventually. In the world we’re entering now, once the bullet train leaves the station, it’s gone. You can’t just sprint and make up the gap later. The pace of change is that fast.
Invest the time and energy now to get onboard. Building AI capability — in your people, your processes, your culture — takes work, time, and investment. It’s not something you can just buy off the shelf. But the risk of missing out, or even just falling slightly behind, is far greater than the discomfort of getting started.
AI isn’t a side project anymore; it’s the new operating system for business. That means as leaders, our job is to create environments that are adaptive, experimental, and always learning. It’s not about knowing every answer — it’s about building organizations that can keep up with the train once it starts moving.
Follow along
You can follow Ashutosh on LinkedIn for more on how he's aligning career paths with AI. And check out Eightfold AI.
More expert interviews to come on People Managing People!