How AI is Transforming Work

Dustin Moskovitz is the co-founder and CEO of Asana and the co-founder of Facebook.

I've been captivated by the potential of AI since my early days in the tech industry. Though the path forward wasn't always clear, I sensed that AI would fundamentally transform the way we work and live.

At Asana, we initially focused on more immediate breakthroughs in machine learning and natural language processing. But I never lost sight of the frontier of AI research, immersing myself in the community by supporting leading AI labs like OpenAI and Anthropic, and working closely with visionaries like Adam D'Angelo, founder & CEO of Poe and an OpenAI and Asana board member.

The emergence of ChatGPT a year ago was a watershed moment. Suddenly, the future we had long envisioned was hurtling towards us at an astonishing pace. We found ourselves at a critical juncture: wait and see how the technology evolved, or dive in headfirst to help shape its trajectory.

For us, there was only one answer. We quickly assembled a dedicated AI team with a clear mandate: find innovative ways to weave the frontier models into the very fabric of Asana in a safe and reliable way. We collaborated closely with leading AI pioneers like Anthropic and OpenAI to accelerate our efforts and stay at the forefront. We launched Asana AI and brought to market powerful generative AI features like smart status updates, smart summaries, and smart answers. The process of taking these first steps unlocked our learning and led us to countless new ideas. We knew this was just the beginning.

Internally, reactions to AI ranged from exhilaration to skepticism. We knew we needed to build literacy and hands-on experience to unite everyone around this transformation. So we launched an internal AI community and immersive workshops, encouraging all employees to tinker with the technology. We stood up Slack channels like #ai-adventures-and-explorations, we started evangelizing the power of AI through new internal use cases, and stories of our personal breakthroughs at company all-hands and in team meetings.

AI for Real Use Cases

Over time, these use cases blossomed in every corner of the company. Our People team built bots for feedback, manager reviews, and self-reflection. Sales & Customer Experience worked on an outbound bot, an Asana product expert, and a call prep bot. Engineering experimented with an API script writer, while Marketing built a content creator and campaign brief bot that was on-message and briefed on our brand design guidelines.

This grassroots exploration sparked individual "aha" moments, which tend to lead to "aha's" for departments. One employee, an SDR at the time, dove deep into designing bots to assist his role. He developed an AI tool that helped him and his colleagues write high-quality, personalized, and well-researched sales emails. Through this process, he became adept at prompt engineering and building AI tools, eventually transitioning to a full-time role focused on accelerating our GTM efforts and driving AI adoption internally. This SDR had no technical background, just initiative and curiosity to get hands-on with AI.

A product marketing manager launched a competitive expert bot that could share his knowledge in a conversational, self-serve manner. The sales team loved that this bot was a true competitive expert, trained on our materials, always available, endlessly patient, and happy to answer any and all questions in a private space. This also freed up the PMM to focus on strategy rather than one-on-one enablement.

Our HR team even built a coaching bot. In one instance, they experimented by recording a 1:1 coaching session with a manager (with the manager's approval) and gave the bot a transcript of the recording, asking it to infer the manager's strengths, weaknesses, and blind spots. The AI's insights formed the basis for a further coaching conversation, with the added benefit that the observations came from an impartial "bot," reducing defensiveness and opening possibilities for learning.

I got the LLM bug myself, tinkering with AI for personal memos, internal communication, and exploring new ways of thinking. Contrary to the popular narrative that using LLMs for writing ends up making you think too little, I found it to be a revelation to see how AI could augment and enhance our cognitive processes. Working with AI helped me write better and think more quickly about strategy, because I could focus on shaping the ideas and spend less time converting them into readable prose.

I often found that if I asked the LLM to critique the work, like identify open questions (and even try to answer them), it would lead me to new ideas, almost by accident. I even tried going back to some older pieces of writing and issuing simple commands like "make the writing quality better" and being amazed by the results.

A Category-Defining Moment

But the moment where it all clicked for me – where it felt like we leaped into the future – was when we brought AI into the flow of collaborative work. I was able to riff with other colleagues, debate ideas, build off each other, and AI was there with us, taking direction, adding perspective, and handling the tedium – it felt like AI was a member of the team. We could even provide feedback as comments in our normal documents editor, and include it for consideration as part of the redraft process.

We began uploading meeting transcripts and sharing Asana task threads as context to the AI so we could collaborate in a very natural way, but still quickly catch up with our "writer" teammate so they can incorporate all the new thinking. The surprising effectiveness of those simple workflows made it clear that the latest AI models are operating at a much higher level than before.

One somewhat unexpected discovery was how this process transformed collaboration with my (human) teammates. Suddenly, receiving feedback and direction felt effortless. I wasn't attached to the words on the page, which I hadn't written in the first place, so I was more open to suggestions. We could have high-level conversations about ideas and structure, without getting bogged down in line edits.

By living this experience, I had more clarity on how humans and AI could work together. We could see the power and potential of integrating AI into Asana's core collaborative features. With our Work Graph data model providing a perfect foundation for AI to understand the complex connections between goals, portfolios, projects, and tasks, we realized AI could work alongside human teammates in a natural way. People can collaborate as they normally would, while AI translates their input into coherent feedback, suggests next steps, and completes assigned tasks. AI is able to advise on what work to do, plan how to do it, and even do some kinds of work assigned to it – all with humans in the driver's seat, and working alongside the AI every step of the way. 

We began to glimpse a future where AI moves beyond isolated chats to embedded, contextual collaboration. Where a marketing team can effectively manage product launches with AIs and humans passing the baton at each step until high-quality deliverables emerge and the product is successfully in market. Or where teams of humans and AI can work together on other typical workflows like resource planning, work intake, and creative production. The technological leap is closer than we thought. We can build this now. Where AI is no longer just a tool, it's a teammate.

Shaping the Future of Work, Together

There's still a lot to figure out in this new way of working. We're grappling earnestly with the immense challenges and open questions around this technology. But we also feel the moral imperative to lean in and actively shape it. And my personal excitement is greater than ever. I've never been more optimistic about our future – the future of work, the impact we can drive for our customers, and the ability for Asana to move closer to achieving our mission of helping humanity thrive by enabling the world's teams to work together effortlessly.

We want to invent the future of work, together, and are eager to share more. We invite you to be the first to know about the latest developments in AI-powered collaborative work here. And if you'll be in the Bay Area, join us at the Work Innovation Summit on June 5th, where we'll be showcasing the future of human-AI teamwork.

Together, let's build a future where every team has the power of AI at their fingertips. Where human creativity and machine intelligence combine to solve the world's greatest challenges. And where work is not just more efficient, but more fulfilling, more impactful, and more profoundly human.

The path ahead won't be smooth. We're going to face some big questions on this journey, and they won't always have simple solutions. We'll have to think from first principles and be willing to throw out some of our old ways of working. But if we reconnect to the deeper meaning and mission that animate our work, I believe we can build something transformative.

So join us on this journey as we work to make technology more empowering, work more purposeful, and collaboration more effortless for all. The future is ours to shape – let's make it extraordinary.

Onwards,

Dustin