Mike Lewis began his career at Scribd, Inc. as an intern more than a decade ago, and never left. Today he leads Scribd and Slideshare, overseeing a combined library of 300M+ pieces of content and 250M users worldwide. 

As AI reshapes how people create, consume, and trust information online, Mike argues that documents aren’t becoming obsolete; they're more important. 

We sat down with Mike to learn about his career trajectory, what he’s learned as the company has grown, and how he sees legacy content playing a role in an AI era.‍

You started at Scribd as an intern and never left. What kept you here? 

I was convinced to leave college early by one of our co-founders. I started in engineering, and have moved through design, strategy, and landed in a sweet spot: product.

What kept me here is the people, and the problem. The company began as a group of smart people trying to make something meaningful. That energy remains.

Scribd, Inc. rewards people who want agency and ownership. If you want more responsibility, you can have it. There's no politics. We just get to make things together.

I have a deep appreciation for the core problem: helping people build knowledge. It's a forever problem.

Our path hasn’t been linear. The original document sharing platform was a rare outlet for people to share their long-form work. Over time, we shifted focus to the premium side of the business, which I ran product for years. 

The user-generated content side was long underrated, but fueling the company. In 2020, I got the chance to return to this part of the business and go deep on it. I ran with it.

You've worked across engineering, product, and strategy. Does that shape how you think about building differently? 

No matter the role, the foundation is relationships. Having worked across functions, I’ve gained a genuine appreciation for everyone's craft – what engineers, designers, and strategists each do. 

Product is about creating the environment for people to do their best work. You're not making the thing, you're enabling others to make it. That's super enjoyable for me. Having lived inside those other disciplines means I can actually do that well. I understand the constraints, creativity, and craft on every side.

You now oversee both Scribd and Slideshare — over 300 million pieces of content, 250 million users. What does that scale actually feel like? 

When I started with the company, our content corpus was in the tens of thousands. It’s now hundreds of millions. 

Over 15 years, a meaningful percentage of the world's population has visited Scribd and gotten something out of it. Tens of millions of people every day use the platform to improve their lives. That's inspiring, but it comes with real responsibility. 

I appreciate that we are helping people all over the world in real ways. From a high school teacher in Peru helping students understand their country's history to someone in the U.S. trying to become a better gardener. 

We play a big role in helping people do their jobs, learn, and find trustworthy information. I don’t take that for granted. 

Documents are often treated as the boring part of the internet, or dated. Why do you think they matter? 

When someone creates a document, they are putting their stamp of approval on it. It’s an act of commitment, with multiple steps: thinking, organizing, writing, sharing. That signals to the reader that a real person spent real time on this, so they can trust it as a well packaged set of ideas. 

It’s not like a tweet or a passing thought. It requires time and effort. 

The democratization of who gets to create them is genuinely exciting. It's not just PHD students or professors; anyone is able to create, and that is beautiful. Someone in Kansas who wrote a guide to building muscle had to put real thought into it – that knowledge deserves to exist and be found.

AI is actually what enables that connection. Beyond helping people craft content,  it helps that content reach the right people. So a person in India searching for muscle-building tips can reference that knowledge.

Documents have prevailed since before paper was invented. They remain the foundation. AI is the connector.

AI can now generate a convincing document about almost anything. Does that change what a "real" document means? 

The world is going through experimentation – everyone can write documents fast. The social construct around documents is shifting, but it hasn't disappeared. 

It used to require significant effort to write something, and readers would give the time back in return (by reading the whole thing). Now it's so easy to put words on a page, but you still need a human to spend a lot of time thinking about it. 

Whether AI assisted or drafted, the expectation is that a human stood behind it, thought it through, and took responsibility. There’s no shame in using AI at various levels, but know that your name is associated with it. 

It’s not that different from the prior world, but AI has made it a lot easier to put thoughts on paper. 

Everyone's talking about AI summarizing content. What happens to the original source?

Summarization is great to quickly connect you to a topic. But to truly learn something requires going deeper. 

We are still humans and we still value connection to other humans. Knowing a real person is the source behind something increases trustworthiness. Original sources actually become more important as the foundation in which AI summaries pull from. 

There can be a harmonious place here. Summaries kick start your journey. The original human-authored document gives you depth that makes you feel ready to bring into your own world.

Legacy institutions (universities, publishers, governments) have traditionally been the gatekeepers of credible documents. Is that still true? 

In the traditional book publishing industry, there’s significant friction associated with contracts and approvals. Who decides which thoughts are worth sharing? Readers actually want to hear from other relatable humans.

Legacy institutions uphold real standards. But they shouldn't be the only type of credible content in the world. 

The growth of user-generated content platforms has allowed more humans to hear how other humans have come through life. A less edited, more human information ecosystem moves us all forward.

There's a narrative that attention spans are shrinking and nobody reads anymore. Do you believe it?

No. Our data has told us people are still reading, many more than ever. But how they read is changing.

The desire to learn, to engage with other people's ideas — that hasn't gone anywhere. What's changed is the volume of what's competing for that attention.

Before AI, you’d stumble across something worth reading once a day. Now you're constantly being shown new things – a colleague's doc, AI summary, platform recommendation. It's a lot to manage.

The behavior – the desire to read, learn, connect – is still there. Humans are going to be the bottleneck here. The world is accelerating how much is put in front of us. It’s a human problem to solve.

What's a problem in the document or knowledge space that nobody is paying enough attention to? 

Burnout and information overload.

AI has made everything hyper-efficient – it's now a waterfall of information where it used to be a trickle. It's disorienting. We are the ones making judgment calls about how to apply all of it to our lives and work. There needs to be products and systems in place to help humans manage it, and that’s been overlooked thus far.

If you could make one wish for a tech advancement, what would it be? 

There have been other trends in the past five years that have led to the potential reduction of human connection (AI and beyond). Without that connection, what's the point?

I’d love for technology to advance human connection, not diminish it. Whatever comes next, I hope it brings us closer together.

Connect with Mike Lewis on LinkedIn.