|
As a partner at Theory Ventures, a VC firm built around deep technology and market research, I spend my days swimming in information: academic papers, market reports, interview notes, and written analyses. Our job is to synthesize these data points into a nuanced perspective to inform our investment decisions. Reading the hype online, its tempting to think you can just delegate anything to AI. But for something so critical to our job, we dont just want it done, we need it to be excellent. How much can AI really do for us? In this piece, I will share: How we structure instructions to get the best analysis out of an AI model Where I critically intervene and rely on my own thoughts How you can get an AI to mirror the way you write When relying on an LLM you often get something that only seems good at first glance: often the AI has missed details, or an important nuance. And for this core part of my job, decent isnt enoughI need output to be excellent. This AI accuracy gap creates a painful cycle where you spin in circles, trying to re-prompt the system to get what you want until youre essentially left rewriting the entire output on your own. In the end, its unclear if AI actually helped at all. The more effective approach is understanding how you (the human) do the thinking and leave writing (i.e., formatting and synthesis) to the LLM. This simple separation is what elevates AI-augmented workflows from decent to exceptional. Heres an example of how we build those kinds of workflows at Theory Ventures, and how you can too. Well illustrate an example with the automation of our internal market research reports. Step 1: Define the thinking process Prepare a document with very detailed instructions on the underlying analysis/construction you seek to achieveclearly outline the context & goals, then dive into all of the details on how you deconstruct a broad analysis: the specific questions you would ask, follow-up sub-questions, how they should be answered with data, and key callouts or exceptions. You can use an AI assistant to help you generate a first draft of this, sharing completed documents and asking it to deconstruct the analysis. But these instructions are critical, so its important to finish writing it by hand and continue to update it over time when you tweak your analysis. Example analysis instructions included in the prompt (note: the full instructions will typically be 2 to 10 or more pages long) Analyze the underlying market structure: Is it fragmented or consolidated? Why? (e.g., high specialization needs, regulatory barriers, network effects, legacy tech debt). How is fragmentation changing over time, and is it different across market segments? Use the following data sources and analyses: . . . Evaluate key market dynamics: What are the typical switching costs? How prevalent is tech debt? What are the typical sales cycles and buyer behaviors? How do incumbents maintain their position (moats)? Use the following data sources and analyses: . . . Step 2: Lay out your human-led analysis Provide your primary analysis, along with raw notes and instructions to the AI. We set our systems up so they require the user to provide their key takeaways and analysis to guide the system towards whats most importanthighlighting areas to focus on, key opportunities, and potential concerns. These are typically four to five detailed bullet points of two to four sentences each. This is the crux of the analysis and should therefore never be AI-generated. Example key takeaways provided to the system: This market has historically been small and fragmented without major software providers. We expect it will grow dramatically, primarily through currently automating labor spend and consolidating a set of point solutions. The underlying demand for this capability will also increase with XYZ challenges. We feel very confident in these two growth levers. Theres substantial concentration at the upper end of the market. Major platforms control around X% of the market and have all invested heavily in their own technology. But below the top-n largest players, there is a healthy cohort of medium-large buyers that have the scale to need this solution but dont want to build it. We think this is sufficient to build a sizeable company, although market concentration and build versus buy remains a key long-term risk. Step 3: Run an interactive Q&A to hone the analysis This dialogue is the most interesting and fun step: Have the system generate questions to clarify the contours of your analysis. Based on the primary analysis, along with the notes and general instructions, the system asks questions about things that either werent clear or had conflicting information/instructions. This helps sharpen the analysis and gives the user the opportunity to share more of their thought process and guidance. Example Q&A: Q from the AI: You said that major platforms have invested heavily in this technology, but conversations with some of those companies indicated an excitement to buy. Do you think that will be common, or were they exceptions? A from the human: Good point. I do think that many of them will buy eventually, but because theyve built a lot of technology internally they are more likely to need a new platform only for certain components, versus buying an end-to-end system. And the very largest companies (top three to five) will build everything in-house. Step 4: Share past work to match tone, not ideas Use previous examples of your work to replicate tone and style only after the scaffolding work is done. Most people skip immediately to this step, but we found (and research shows) that providing finished examples is most useful just to match tone and writing style, as opposed to shaping the analysis itself. In researching the best AI-native products, weve seen that practically all of the work goes into defining the thinking and analysis portion of the problemdetailed instructions, guidelines, orchestration, and toolingso the AI system knows what it should do and just executes on it. At Theory Ventures, weve started to mirror the same system by developing highly-constrained, human-in-the-loop workflows that direct the analysis, leaving the LLM to execute basic information extraction and synthesis. Thats how weand our AI systemshave started working smarter. Not by asking AI to think for us, but by helping it think better.
Category:
E-Commerce
For decades, the plastic crisis has always felt far away, whether through time or across distance. But unfortunately, were no longer talking about environmental pollution out there in the ocean. Microplastics, tiny fragments that come from the breakdown of everyday plastic items, are now inside all of us, turning this from a theoretical risk into a shockingand deeply personal (physically) reality. Despite this fact, most Americans remain unaware of just how prevalent microplastics are in our lives. New national research that we conducted with our partners at The 5 Gyres Institute paints a troubling picture: while 77% of Americans say theyve heard the term microplastics, only 49% actually understand what it means. Even about half of people51%know its often a result of larger plastic breaking down. The knowledge gap That knowledge gap is more than an academic concern. Its a public health crisis, especially when you consider that, after the term is defined for survey respondents, 90% of Americans state that theyre worried about microplastics in the human bodyand theyre right to be. Science confirms that these particles have been found in breast milk, placental tissue, lungs, brains, blood, and more. And studies are increasingly linking microplastics to serious health impacts, including cancer, heart disease, hormone disruption, and infertility. But even when you close that knowledge gap, people who care often feel stuck. Our research shows that 70% of Americans dont know how to reduce their exposure to microplastics and 67% cant name a single company actively working on the problem (were hoping to change that!). That sense of powerlessness is as dangerous as the plastic itself, because people want better. They just dont know where to turn. At Grove, weve seen firsthand that Americans are searching for answers and theyre looking to us: to companies, brands, and private-sector leaders. They want healthier homes, safer products, and more sustainable choices. They want corporations to leadnot with vague promises, but with bold, measurable action. This is our collective moment. A solution Consumers didnt create the plastic crisis. We, the private sector, did. For decades, our industries have driven plastic adoption in product design, packaging, and sourcing. And we were lied to and manipulated by the petrochemical and plastics industry that shaped this system. Now, we, the private sector in 2025, must dismantle it. That means going beyond plastic. It means rejecting outdated systems that rely on single-use packaging, microbeads, and petrochemical-based materials. It means investing in compostable and refillable formats, shifting supply chains, being transparent about ingredients and sourcing, and leaning into the circular economy. It means learning and being aware of the impact plastics are having in our bodies and environments. It means supporting legislation, like the newly introduced bipartisan Microplastics Safety Act, which calls on the FDA and HHS to investigate and report on the health impacts of microplastics. Most importantly, it means refusing to offload responsibility onto consumers and admitting that recycling, long touted as a solution, simply isnt enough. Only 5% of plastic is recycled and the rest ends up in landfills, incinerators, or breaks down into microplastic particles that pollute our air, food, water, and (if not abundantly clear by now) our bodies. At Grove, we remain unwavering in our commitment to eliminate plastic from the products we make and selland to empower others to do the same. But we cant do it alone. The cost of inaction Consumers are demanding accountability. Our research shows that 79% of Americans believe microplastics represent a human and environmental emergency; 82% believe companies should be doing more. But only half (54%) believe that businesses are actually stepping up. That gap is where trust and long-term relevance will be won or lost. The cost of inaction is rising. Not just in terms of public health, but in trust, consumer confidence, and regulatory risk. There will come a time soon when inaction on microplastics will be seen for what it is: negligent at best, and reckless at worst. Companies that continue to delay action on plastic pollution arent just making a business decision. Theyre making a decision that directly impacts human health. Brands that cling to plastic-heavy models are effectively choosing profits over people, and theyll have to live with the consequences. But brands that choose to lead? Theyll be rewarded with consumer loyalty, resilience, and relevance in a world thats rapidly waking up to this crisis. The science is clear. The public is paying attention. The future will not be plastic. And the time for action is now. Jeff Yurcisin is CEO of Grove Collaborative.
Category:
E-Commerce
As U.S. climate policy was noisily dismantled in Washington over the spring and summer, another climate story unfoldedquieter, faster, and broadcast to millions. It unfolded in the streets of Monaco. So Paulo. Shanghai. In the form of all-electric race cars tearing through city centerscheered on by fans living the transition to a low-carbon world, not waiting for it. Formula E: A global entertainment platform Launched just over a decade ago, Formula E now reaches half a billion fansmany of whom are new to motorsports. Not because it promised sustainability. But because it delivered a better product: short, high-drama races. Urban venues. A streaming-ready format. Cultural relevance in an EV-first world. Its a playbook worth studying for any company trying to bring climate innovation to the mainstream. This isnt about messaging. Its about strategy. From clunky to cutting edge When Roger Griffiths first saw a Formula E car in 2014, he wasnt impressed. A veteran of IndyCar, Le Mans, and Formula 1, he knew a lot about going fast. And this wasnt it. The battery was huge, heavy, and underpowered. The performance? Underwhelming. But Formula E wasnt starting from scratch. It was pulling from the top shelf of global motorsport. What struck him wasnt the hardware. It was the names showing up anyway: Michael Andretti, Alain Prost, and Frank Williamslegends who had built dynasties in IndyCar and Formula 1. Even Richard Branson had joined the grid. We cant afford (for) this to fail, Griffiths recalled on the Supercool podcast. Too many people have too much invested. Formula E didnt begin with speed or range. It started with credibility. And in the early days of climate tech, that buys you time to iterate toward something better. So they did. Designed for a new kind of fan Formula E didnt mimic Formula 1. It built a motorsport tuned to a new era. Races last just 45 minutestight enough for modern attention spans, long enough to create drama. The circuits run through the hearts of global cities, not remote tracks. Fans take public transit or Uber to races. The vibe? Less pilgrimage, more pop-up festival. The audience is younger, urban, and digitally native. Many arent interested in owning a car at all. Young people today dont necessarily want to own cars, said Griffiths. Were catering to a crowd that thinks differently about mobility. Formula E recognizes that. Meanwhile, the technology caught upfast. Jaguar used race-day insights to improve the range of its I-PACE SUV. BMW co-developed systems between i3 engineers and race teams. Formula E became a proving groundnot just for fans, but for the EV industry. Built to evolve Unlike legacy motorsports, Formula E gave itself permission to break with tradition. It experimented early and often: Fan Boost. Attack Mode. Interactive features lifted from gaming culture. Some flopped. Others stuck. But the league kept shipping, learning, and moving forward. The old me wouldve said, What a stupid idea, Griffiths said of Attack Mode, which gives drivers a temporary power boost if they hit a marked zone on the track. The new me said, Im not surebut Ill give it a go. Formula E doesnt wait for perfect. It tests ideas in publicon race day, with millions watching. Either way, the race goes on. The sport gets better. That mindset isnt just tolerated, its structural. Formula Es governance enables change. Its culture rewards it. 5 lessons for climate innovators Innovators can learn these five lessons from Formula E. 1. Turn constraints into strengths.Early EVs couldnt finish a full race. Formula E shortened them to 45 minutes, creating tighter, more intense competition perfectly tuned for social media highlight reels and streaming. 2. Design for urban lifestyles.Electric cars are quiet enough to race in city centers. Fans dont need to drive. They grab an Uber and plug into the experience as part of modern life. 3. Iterate in public.Formula E doesnt hide experiments. It ships them in real time, where fans become part of the process. Innovation is part of the show. 4. Let climate be the platform, not the pitch.Sustainability underpins the whole thing. Sponsors dont need convincing. Fans dont feel preached to. Thats what makes it scale. 5. Design for whats emerging.Formula E didnt retrofit old formats for electric race cars. It aligned with a modern, urban culture: streaming-first viewing and shared mobility. These behaviors define where were heading. A better future, built for speed Formula E didnt scale by talking about emissions. It scaled by delivering an incredible fan experience. It understood how younger audiences live, move, and engageand built a sport around that. It made the low-carbon future feel inevitable, not through fear, but through energy and excitement. And it proved something essential: Climate innovation doesnt have to trade performance for principle. It doesnt have to trade anything at all. Josh Dorfman is CEO and host of Supercool.
Category:
E-Commerce
All news |
||||||||||||||||||
|