Drivers of Innovation

Should you try Google’s famous “20% time” experiment to encourage innovation? We tried this at Duolingo years ago. It didn’t work. It wasn’t enough time for people to start meaningful projects, and very few people took advantage of it because the framework was pretty vague. I knew there had to be other ways to drive innovation at the company. So, here are 3 other initiatives we’ve tried, what we’ve learned from each, and what we're going to try next. 💡 Innovation Awards: Annual recognition for those who move the needle with boundary-pushing projects. The upside: These awards make our commitment to innovation clear, and offer a well-deserved incentive to those who have done remarkable work. The downside: It’s given to individuals, but we want to incentivize team work. What’s more, it’s not necessarily a framework for coming up with the next big thing. 💻 Hackathon: This is a good framework, and lots of companies do it. Everyone (not just engineers) can take two days to collaborate on and present anything that excites them, as long as it advances our mission or addresses a key business need. The upside: Some of our biggest features grew out of hackathon projects, from the Duolingo English Test (born at our first hackathon in 2013) to our avatar builder. The downside: Other than the time/resource constraint, projects rarely align with our current priorities. The ones that take off hit the elusive combo of right time + a problem that no other team could tackle. 💥 Special Projects: Knowing that ideal equation, we started a new program for fostering innovation, playfully dubbed DARPA (Duolingo Advanced Research Project Agency). The idea: anyone can pitch an idea at any time. If they get consensus on it and if it’s not in the purview of another team, a cross-functional group is formed to bring the project to fruition. The most creative work tends to happen when a problem is not in the clear purview of a particular team; this program creates a path for bringing these kinds of interdisciplinary ideas to life. Our Duo and Lily mascot suits (featured often on our social accounts) came from this, as did our Duo plushie and the merch store. (And if this photo doesn't show why we needed to innovate for new suits, I don't know what will!) The biggest challenge: figuring out how to transition ownership of a successful project after the strike team’s work is done. 👀 What’s next? We’re working on a program that proactively identifies big picture, unassigned problems that we haven’t figured out yet and then incentivizes people to create proposals for solving them. How that will work is still to be determined, but we know there is a lot of fertile ground for it to take root. How does your company create an environment of creativity that encourages true innovation? I'm interested to hear what's worked for you, so please feel free to share in the comments! #duolingo #innovation #hackathon #creativity #bigideas

94% of the time, engagement with AI (including Generative AI) leads to asking a wider range and variety of questions than the user would have otherwise. This says it all. Just the way we learn so much from others by just talking to them, the power of GenAI is to stimulate our curiosity and lead us to explore ideas and possible solutions we may not have considered otherwise. That’s how we achieve greater personal performance. Today, PMI released a report on prompt engineering for project management. It's here: https://lnkd.in/eBZDsy9j Prompt engineering really is a new discipline in itself. PMI's goal is to help project professionals engage with GenAI tools effectively and efficiently so they can get the most out of their interactions. The report explores how to better use GenAI tools, the ethics of using them, and it offers strategies and techniques for holding a conversation with a GenAI tool. You’ll rarely get the perfect result after only one question... but if you use the right prompts and techniques, you can get better results, faster. As we have all experienced, GenAI can actually lead to more innovative solutions than you might expect. Please check out the report, let me know what you think, and please repost, so we keep helping the whole profession together. Feel free to share your favorite tips or prompts below as well.   Personally I like this quote from Bart Gerardi, because it fits with my own observations of how both Chat GPT and Project Managers tend to operate: “Nobody wants a quarter-inch drill. What they want is a quarter-inch hole. We’re so accustomed to asking for the drill, but we’re not accustomed to asking for the actual outcome that we want. To use GenAI properly, we need a shift in how project professionals think. They should be asking for their outcomes.” Project Management Institute #ArtificialIntelligence #AI #ProjectManagement #PromptEngineering"

India’s IITs are world-class for undergrads—so why do they lag in research & postgraduate excellence? Despite immense talent and bright spots (IISc, select IIT labs), India’s research impact trails Western and leading Asian peers (China, South Korea, Japan, Hong Kong, Singapore). Closing this gap is essential for the innovation economy. What’s holding India back (plainly): Thin research funding. R&D ~0.7% of GDP; labs and core facilities underfunded. Skewed incentives. “Publish a lot” beats “publish well,” hurting citation impact and reputation. Talent leakage. Low/irregular PhD stipends and weak early-career paths push top students abroad. Low internationalization. Too few foreign faculty/students; limited international co-authorship. Teaching legacy. Institutes built for UG excellence; research pivot is recent and uneven. Red tape. Centralized rules slow hiring, procurement, partnerships, new programs. Shallow industry links. Limited sponsored research and tech transfer impede translation. Seven moves to bend the curve: 1) Fund like it matters. Lift R&D toward 2% of GDP; ring-fence facilities/core labs. Grant financial autonomy (endowments, flexible fees for specialised/intl programmes, retain IP income). 2) Reward excellence, not volume. Promotions/hiring weighted to field-leading work (top-tier journals, patents/products). Zero tolerance for predatory outlets. Time and seed grants for bold ideas. 3) Make India the best PhD destination. Raise/index fellowships; pay on time. Expand postdocs, mentoring, early-career grants. Create return-chairs for diaspora; enable globally competitive offers. 4) Go global by design. Target 20%+ foreign faculty in priority areas; scholarships for international PGs; joint PhD academies; MoUs tied to co-advised theses, shared labs, mobility. 5) Back fundamentals and translation. Fund blue-sky science alongside mission centres. Professionalize TTOs, cover patent costs, build incubators integrated with graduate training. 6) Govern for speed and accountability. Real academic/administrative autonomy with transparent dashboards (research quality, time-to-procure, graduation, placement, spin-outs). Regulators as facilitators. 7) Make industry a co-investor. Matching grants/tax credits for sponsored research; embedded industry labs; normalized faculty R&D sabbaticals; industry adjuncts; credit-bearing student build-sprints. A pragmatic horizon: With aligned money, metrics, mobility, and management, IITs/IISc/IISERs and leading universities can move from “strong teaching, patchy research” to globally competitive research engines within a decade. This is less about rankings, more about an innovation flywheel: attract talent → fund excellence → produce breakthroughs → translate at scale → reinvest. #HigherEd #Research #Innovation #India #IIT #IISc #GradSchool #STEM #Policy #Talent #RandD #UniversityReform #DeepTech #MakeInIndia #BrainGain #AcademicExcellence

PhD students often ask me: What should I work on given academia’s compute constraints? 🤔 My answer: Focus on the questions only fundamental research can solve. Industry optimizes for short-term gains. Academia thrives where the unknown is too risky, too slow, or too disruptive for companies to explore. Some ideas: 🔹 Why do LLMs have reasoning capabilities? Scaling laws don’t explain everything—what are we missing? 🔹 Beyond task performance: Design benchmarks that capture how models behave, not just how well they score. 🔹 Disentangling learning from data curation: What happens when we separate model capabilities from dataset artifacts? 🔹 Prototype what industry won’t risk: Multi-agent AI foundations, energy-efficient training, or alternative compute paradigms. The goal isn’t to chase SOTA but to interrogate the unknown. The best work in fundamental AI research might take years to mature—but it’s what redefines the field. 💡 What fundamental AI questions do you think academia should tackle? Let’s discuss. 👇

Innovation isn’t just about new products. It’s about how you structure, deliver, and capture value—across your entire business model. In their book, "Ten Types of Innovation" (2013), Keeley et al. outline a powerful framework outlining no less then 10 types of innovation: Configuration 1. Profit Model – How you make money 2. Network – How you collaborate 3. Structure – How you organize 4. Process – How you operate Offering 5. Product Performance – What you offer 6. Product System – How offerings work together Experience 7. Service – How you support users 8. Channel – How you deliver value 9. Brand – How you're perceived 10. Customer Engagement – How you foster loyalty Most innovation efforts focus narrowly on the product. But real advantage comes from orchestrating multiple innovation types, often in combination. If you're looking for new strategic levers, this framework is a great place to start. Which of the ten are you already investing in?

Many organizations approach innovation the way they approach budgeting or operations. They create roadmaps, timelines, and committees designed to produce breakthroughs on schedule. But the history of technology suggests something different. Most meaningful innovations do not arrive neatly on a calendar. They appear unexpectedly. A new idea. A technical breakthrough. A surprising connection between two things that previously seemed unrelated. The real challenge is making sure your organization is ready when those moments appear. The companies and institutions that consistently innovate tend to invest early in talent and technical capability. They build cultures where experimentation is encouraged and where people are willing to test new ideas. They maintain the flexibility to pursue unexpected opportunities and move quickly when promising ideas appear. Innovation rarely begins as a fully formed plan. More often it begins as a possibility that only a few people recognize at first. The advantage goes to the organizations that have prepared themselves to recognize that moment and act on it. You may not be able to schedule inspiration. But you can build teams, systems, and cultures that are ready when it shows up. #SchmidtSights

💡 New Zealand recently defunded humanities and social science research. Canada is doubling down on it to drive innovation. Who’s right? New Zealand’s government recently cut funding for social sciences and humanities (SSH) research, arguing that science and technology - not SSH - are the key to solving our productivity and growth challenges. Minister Judith Collins stated that only "core sciences" like physics, chemistry, and engineering will deliver real economic impact. Meanwhile, Canada is making the opposite bet. President of Canada's Social Sciences and Humanities Research Council (SSHRC), Ted Hewitt, argues that SSH is essential for turning science and technology into innovation and productivity gains. In this article, he lays out why: ✅ Innovation is more than invention. Science and technology create new products, but SSH research ensures they are adopted, commercialized, and integrated into society. Without SSH insights, many breakthroughs never reach their full potential. ✅ Industry needs SSH expertise. Canadian researchers are collaborating with businesses—from airlines balancing sustainability and profitability to fintech companies designing more inclusive products. SSH research is actively shaping business success. ✅ Workforce creativity fuels economic growth. The World Economic Forum’s Future of Jobs report lists creativity, critical thinking, and problem-solving as top skills for the future—precisely the skills fostered by SSH education. ✅ Better policy means better productivity. SSH research provides data-driven insights that help businesses and governments design smarter regulations, improve economic strategies, and remove barriers to growth. If New Zealand wants to lift its productivity and drive innovation, cutting SSH funding might be the worst decision it could make. Canada recognizes that science alone doesn’t drive economic success. It's the integration of SSH that makes innovation work. With Shane Reti set to take up the new universities portfolio plus Science, Innovation, and Technology, it's time for a rethink. #SocialSciences #Humanities #Marsden #Universities #Research #Productivity #NewZealand

The hardest step in Goldratt's Theory of Constraints isn't finding the bottleneck. It's step three: subordinate everything else to the constraint. Translation for software teams: if your code review process can absorb 1.2x the current volume, don't generate 2x. If you can't measure outcomes on twice the features, don't commit twice the features to production. This means deliberately throttling AI output. I'll say that again because it sounds heretical. It means looking at a tool that can generate code faster than ever, and choosing to slow it down. Not because the tool is bad. Because the system downstream can't absorb what the tool produces. Most organizations do the opposite. They celebrate the increased commit volume. They trumpet the PR throughput numbers. They showcase the individual productivity gains. Meanwhile, review queues grow. CI recovery time gets worse. Deployment problems increase. 96% of the most frequent AI users end up working evenings and weekends. They optimized the part of the system that wasn't the bottleneck. That's not progress. That's inventory accumulation. The question isn't "how much code can we generate?" It's "how much code can our system absorb, validate, and deliver?" Generate up to that limit. Not beyond it. If that feels wasteful, you've identified the real investment priority: increasing the capacity of the constraint. But exploit and subordinate come first. Throwing money at capacity before you've maximized what you already have is just expensive inventory management. #TheoryOfConstraints #EngineeringLeadership #AIReadiness

Generative AI (GenAI) has ushered in a renaissance age for the generalist. For years, organizations have spent a disproportionate amount of capital hiring hyper specialized talent with deep technical knowledge. Now, with the democratization of #GenAI, the value offered by hiring ‘capable generalists’ is on the rise. People who articulately frame their thoughts, pose well-formed questions (prompts), and exercise #AI tools to their advantage, stand to benefit greatly. The demand for specialized AI talent - model developers, AI ops talent, and engineers to build and maintain infrastructure - will persist. But demand for non-technical talent is shifting to a more balanced state. Those who have the skills to extract value from platforms are becoming as valuable to organizations as those who build them.  I strongly encourage business leaders to incorporate skills like curiosity, critical thinking, and effective writing into their hiring profiles. These skills are becoming increasingly important - and valuable - in this next phase of technology and operations.

Creativity is up. Diversity is down. Everyone’s talking about how GenAI speeds up creativity. But speed isn’t the real shift. What’s actually changed is how we create, and who does what. A few years ago, being creative meant starting from zero. Now it means knowing what to keep, what to cut, and what to remix. The best creatives I know aren’t just creators. They’re: ➡️ Curators, picking from AI outputs ➡️ Editors, tightening what matters ➡️ Prompt writers, shaping the next round GenAI doesn’t replace the process. It reshapes it. Taste matters more than originality. Workflow matters more than perfection. But here’s the catch: When everyone uses the same tools in the same way, The work starts to look the same. Sameness scales. A Science Advances study showed that while GenAI boosts individual creativity, it reduces the diversity of what gets made. Polished? Yes. Different? Not always. If we’re not careful, AI might make us more productive but less original. So the next creative skill isn’t generation. It’s direction. → Inject human unpredictability → Break the template → Ask better questions than the model can answer Curious: What’s one way you’re protecting originality in your creative or product work? Let’s trade notes 👇 🖼️ Image: AI-generated via Ideogram