Rohan Chandran

the bets are all in on ai Essentially, I think the story of these circular deals is really no different than the big macro story about the AI boom itself. AI is the most promising technology out there, and we’re simply betting a lot of our economy on the hope that we can exploit its full potential. Should we worry about AI's circular deals? - by Noah Smith

3

It's like asking your landlord to also be your tenant rights lawyer. The governance layer your agents need should never be owned by the same company selling you the inference. This is exactly why we built ContextGate to be model-agnostic from day one. People ask us all the time: "What happens if Anthropic or OpenAI builds what you're building?" They won't. And this story is the proof. These companies are in the business of selling you their model. Their incentive is to lock you into their ecosystem, their pricing, their infrastructure. Letting you seamlessly swap to a competitor's LLM when they jack up prices? That will never be on their roadmap. It's the opposite of their business model. This founder's entire product broke in 13 minutes because she was dependent on a single vendor's pricing decision. One email at 11:47. Product dead by 12:01. That's not a bug. That's the architecture working exactly as the vendor designed it. At #ContextGate, every agent call goes through one API endpoint. You can route to Claude, GPT, DeepSeek, Qwen, Llama — whatever makes sense for the task and the cost. Swap models with zero code changes. If one provider changes their pricing on a Friday afternoon, your product keeps running on Monday morning. One API. Any model. No lock-in. That's not just a feature — it's the whole point.

3

Had a nice discussion with Allie Garfinkle from Fortune yesterday (thanks Allie!). TL;dr - while #VCfunding for pure #ecommerce is clearly down (#D2C pullback, trade wars, etc.), #AI for #commerceinfrastructure is looking strong. The consumer ecosystem has always been an early adopter of technology. And the solutions that work in retail are often able to serve other verticals with large audiences (financial services, media, home services, etc.). Success in the large enterprise is especially transferable. That's what we're seeing across categories, from Auxia to Lyric. And it's a big part of our newest investment (announcing soon!) https://lnkd.in/gBxH_atC

34

1 Comment

We're often asked how Grindr is prioritizing user privacy as we become an AI-first company. It’s a great question and one we spend a lot of time thinking about. Our Chief Product Officer, AJ Balance, outlines our approach in this post 👉https://lnkd.in/gxZ37T_7 #Grindr #privacy #AIinWork

24

2 Comments

AI is changing the structure of most companies we see. Lots more coming from us on how startups can build faster, leaner teams -- and where we see all of this going. Great piece from Pete Flint to kick it off: https://lnkd.in/dcUebcJj

57

1 Comment

This week, I’m taking a deeper dive into META. https://lnkd.in/gu-w5cq6 Trailing twelve-month (TTM) revenue through 9/30/25 was $189 billion with a 43% operating margin. The Wall Street consensus is for revenue to increase 21% in 2025 to $199 billion and 18% in 2026 to $235 billion. 99% of revenue and 122% of operating income comes from the Family of Apps division. This is comprised of Facebook, Instagram, WhatsApp, etc. The Reality Labs division generated $2 billion of TTM revenue with an operating loss of $18 billion. Value Score META generates 61 Value Points from 43% TTM operating margin plus 18% projected revenue growth. Dividing this by the operating multiple of 17 provides a Value Score of 3.5. This puts META firmly in Buy territory. There was a lot of concern at the last earnings call when META mentioned they were going to invest heavily in AI. Given that statement, I wanted to look at the free cash flow of the company. TTM capital expenditures were $63 billion, 68% higher than $37 billion for 2024. The company’s cap ex guidance for 2025 was $70-72 billion. For 2026, I assumed the cap ex would grow slightly from the 2025 guidance. This created a multiple of 42 for the enterprise value divided by the projected unlevered free cash flow. The adjusted Value Score of 1.5 is firmly in Hold territory. Segment Valuation Another technique for valuing the company was to apply revenue multiples to each division and see if the projected equity value is greater than or less than the current market value. The median trailing revenue multiple for companies with a market value greater than $100 billion is 3.36, while the median Value Points is 25.84. This implies that each 7.7 Value Points is worth a 1x multiple. For simplicity and to be generous, I assumed a 1x multiple for each 7 Value Points of the division. I applied an 11x multiple to the Family of Apps’ $187 billion of revenue. I was very generous in applying a 3x multiple to Reality Labs’ $2 billion of revenue, which is really a startup valuation metric rather than a public company metric. The formula suggests that META has 27% potential upside in its stock price. As always, please do your own research before making any investment decision. Read this and other posts about public stocks on my Substack:  https://lnkd.in/g646RNnS

4

AI startup AGI infinite revenue involves tech bros lying about it like Chungin Lee and Cluely - foreshadowing scope of incompetent ARR @ in this case Andreessen Horowitz and Shaheer Sandhu w Susa Ventures et al - story below, thx to Julie Bort w good journalism

4

1 Comment

**Navigating Leadership in AI-Driven Platforms** Meta's recent appointment of Connor Hayes as the head of Threads exemplifies the strategic importance of generative AI in shaping user engagement. This move not only enhances their product offerings but signals a shift in how social platforms will utilize AI to create richer interactions. At ElevenX Capital, we believe that understanding such trends is critical for Limited Partners as they assess investment potentials in technology-driven ventures. How do you foresee generative AI influencing the social media landscape in the coming years? #investing #innovation #venturecapital #entrepreneurship

3

How do you decide which cognitive capabilities to preserve and which to let go? I've been developing what I call the Cognitive Portfolio Framework. 𝗧𝘄𝗼 𝗱𝗶𝗺𝗲𝗻𝘀𝗶𝗼𝗻𝘀: ▶️Is this capability core to your identity, or purely functional? ▶️Should you preserve it, or is delegation appropriate? 𝗙𝗼𝘂𝗿 𝗾𝘂𝗮𝗱𝗿𝗮𝗻𝘁𝘀: 🏛️ 𝗦𝗮𝗰𝗿𝗲𝗱 𝗖𝗮𝗽𝗮𝗯𝗶𝗹𝗶𝘁𝗶𝗲𝘀 (𝗜𝗱𝗲𝗻𝘁𝗶𝘁𝘆 + 𝗣𝗿𝗲𝘀𝗲𝗿𝘃𝗲) - Protect these. Practice them regularly, even when AI could handle them. 🧰 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗰 𝗥𝗲𝘀𝗲𝗿𝘃𝗲𝘀 (𝗙𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝗮𝗹 + 𝗣𝗿𝗲𝘀𝗲𝗿𝘃𝗲) - Maintain for contexts where AI isn't available. Resilience, not identity. 🦋 𝗜𝗱𝗲𝗻𝘁𝗶𝘁𝘆 ��𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 (𝗜𝗱𝗲𝗻𝘁𝗶𝘁𝘆 + 𝗗𝗲𝗹𝗲𝗴𝗮𝘁𝗲) - Capabilities you're choosing to release. Conscious evolution. 📤 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝗗𝗲𝗹𝗲𝗴𝗮𝘁𝗶𝗼𝗻 (𝗙𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝗮𝗹 + 𝗗𝗲𝗹𝗲𝗴𝗮𝘁𝗲) - Let these atrophy. Reclaim the cognitive space. The goal isn't to preserve everything. The goal is to choose — consciously managing what you keep and what you trade. Full framework in this week's article: https://lnkd.in/g4SUYYAV

2

1 Comment

Here's the brief moment of self-reflection we took last weekend from deep inside the absolute center of the pulsing heart of the AI movement. The Curve at Lighthaven in Berkeley was probably the most important gathering I've been to in well over a year. Last year Kevin Roose called it "an event where history was happening" and I understand now what he meant -- it's the only place I've seen where 250 senior leaders from Google, Anthropic, and OpenAI, leading economists, researchers, congresspeople, and social sector leaders get together to butt heads and learn about each others' biggest bets and hottest takes. Have you asked yourself recently what your P(Doom) is? I'll be sharing more about what I heard in other formats, but a few highlights: - It's sometimes dizzying to see senior leaders from the big 3 AI Foundries agreeing with each other on so much, slapping each other on the back, and cracking friendly jokes with each other, while simultaneously expressing deeply-held skepticism about cooperation being a solution for their (shared!) safety concerns. - AI Doomers and e/accs tend to have a ton in common, but often differ on 1 or 2 very specific assumptions. - I saw a fascinating meeting of minds between pragmatic economists and fast-takeoff-safety doomers. I think I saw them start to understand each other better, and what at the core is the fundamental difference of perspective. - The Curve will definitely show up in the history books when (surely countless) retellings of this time start to be published. - Extraordinary robotics demos that reset my understanding of where we are on that curve. - Having social sector and globally-minded voices in the room was essential to the right conversations happening: not just about what's possible, but how all aspects of society will likely react. - One thing is for sure: We live in the most interesting of interesting times. Exciting, to say the least. Extraordinary work by Golden Gate Institute for AI bringing it together, and kudos to Rachel Weinberg Steve Newman and Taren Stinebrickner-Kauffman for hosting this absolutely *essential* summit. Great as always to see Yoshua Bengio, Aaron "Ronnie" Chatterji, The Collective Intelligence Project team, Kevin Barenblat, Tim O'Reilly, and the many other brilliant and generous speakers and guests. LinkedIn won't yet let me combine photos and videos together in a post, so I'll put some photos in the comments.

33

6 Comments

Verizon’s CEO Dan Schulman said it loud and clear: “A strategic approach that relies too much on price without subscriber growth is not a sustainable strategy.” And even more directly: “You have to treat people like humans, not like accounts.” In other words: You cannot increase price without increasing value. This isn’t a pricing problem. It’s a product operating model problem. For years, the playbook was simple: • Charge more • Bundle more • Hope customers stay Meanwhile: • Customer value declines • Differentiation is shrinking, switching costs are lower than ever • Churn increases: Verizon lost 2M+ customers tied to pricing decisions That’s not a pricing issue. That’s the gap between what customers pay and what they get. Network quality and legacy reputation alone are no longer enough. You have to deliver value at every step of the customer journey. Not just at signup. LTV isn’t captured at the point of sale. It’s built into the daily experience. That requires a different approach: • Focus product and operational teams on what really matters to the customer • Pricing should match the value delivered • Keep it simple: clear plans, clear value, easy to understand “Treat people like humans, not accounts” sounds obvious. Here is where most companies fail. A fundamental shift: ✅ Revenue-first → Value-first ✅ Transactions → Relationships #ProductStrategy #CustomerExperience #Retention #GTM #ProductLeadership #LTV #Verizon

5

3 Comments

Important to remember that the public voices most vocally prophesizing mass societal disruption from AI happen to be selling… AI. Only industry I can think of (other than big Tobacco) that goes out of its way to advertise the potential for negative impacts from the work they do. Obviously serves their interests to reinforce their own importance and perpetuate the current hype cycle. But also shapes a narrative that they are in no way responsible for the consequences of their actions and that it is someone else’s job to clean up after it. It’d feel a lot less self-serving if they were bringing proposals or solutions into the public debate, instead of just highlighting risks.

52

3 Comments

Sierra landed at No. 6 on the 2026 CNBC Disruptor 50, up from No. 48 last year. Anthropic and OpenAI took the top two spots. Worth pausing on what that ordering says about the AI stack. The foundation labs are building the engines. Sierra is one of the companies turning those engines on for businesses to run on agents. For companies like ADT, Rocket Mortgage, GAP, SiriusXM, Cigna, Nordstrom, and Singtel, these are AI agents resolving mortgage refis, processing claims, handling returns, managing subscriptions. Millions of interactions a month. Real money, real customers, real outcomes. The model layer and the application layer are not the same race. Both are massive. Both will produce category-defining companies. And the application layer is where most of the value will land for the enterprises buying this technology. Excited to keep building toward that with our customers. If you're thinking through where agents go next in your business, happy to compare notes. https://lnkd.in/emjVWVgX

86

Failure serves as our greatest teacher, offering invaluable lessons that can lead to building a better company. While some failures are massive in scope, others may be smaller and less catastrophic, yet each provides critical insights. The Jawbone story stands out as one of the most significant failure narratives, rich with lessons that can be applied to future endeavors. Here's my latest substack.... https://lnkd.in/gp5-dgv9

13

4 Comments

Excited to share that Guide AI will be meeting with Athena Alliance leaders in the EPD space to share how to effectively build with AI. I'll be sharing the learnings between v0 of our product to our now enterprise-grade product that was bolstered by bringing on an experienced leader like Bhaidaya Baker. I'll also be sharing where I think AI is going and what that means for founders building for the highly compliant, highly regulated space. Sign up below! https://lnkd.in/e5NZ5qZy

19

1 Comment

Talking to users and thoughtfully integrating their needs and wants was woven through everything we did at IDEO—threaded through every prototype and product decision. Reading my colleagues Joanne and Leo's latest work brought that to mind. They describe a future where research runs continuously, with AI agents recruiting participants and instantly surfacing insights. Synthetic personas, carefully trained on consumer signals, mirror human behavior closely enough to extend live studies and compress timelines from weeks to hours. Today user research mostly runs in fits and starts, prohibitively expensive and locked away in dedicated project cycles when it could be a constant companion, seamlessly integrated into the product development process. Joanne and Leo lay out the full vision here: https://lnkd.in/gEUBSG_J

35

3 Comments

I've had the pleasure of connecting and working with fantastic marketplace founders in recent years, mostly via the Everything Marketplaces community. The ways that marketplaces are built and marketed are changing rapidly, but the best teams are seeing these changes as opportunities and not setbacks. Here's what a few are doing exceptionally well: Jawa.gg – Amanda Stefan and team have built a textbook example of community-led growth for P2P marketplaces. Their commitment to nurturing the gaming community through channels like Discord and influencer partnerships drives meaningful traffic, feedback, and genuine excitement to their PC gaming builds and parts marketplace. Rebound Jerseys – Aidan Scurato and team are laser-focused on hockey jerseys as Rebound's beachhead into the wider world of sports collectibles. This tight focus sharpens prioritization and allows them to build things their target community loves. For example, their collection logging feature creates a pathway for people to engage with their site even when they're not actively buying or selling. Stock: The Logistics of Inventory Donation – Abby Nawrocki built a platform connecting brands and 3PLs with nonprofits to simplify donating excess inventory. I've never encountered a marketplace quite like this, and it's clear to me how much Abby and team are driven by their mission and vision. I cannot wait to see the positive impact this company is going to make. Athletes Untapped – AU is a marketplace to connect vetted sports coaches with athletes, and talking to Gene Williams recently, it was clear to me how much his team cares about building an amazing tool for this community. Their foundational SEO work has also laid the groundwork for organic growth in the years ahead. Conquer Pickleball – I have immense respect for the way that Louis Long and team followed the data and user feedback to focus exclusively on pickleball, helping players of that sport find local games and courts. That disciplined focus has created an awesome app for the community and is a great example of how narrowing your scope can lead to outstanding product outcomes. ThrowbackBuys – While many used goods marketplaces focus on a niche category or feature set, LeMarc J. built his marketplace around an emotion: nostalgia. Talking to LeMarc, you hear his obsession with the psychology behind why people buy collectibles, which is a super intriguing way to approach marketplace. These are just a few examples of marketplaces I'm excited about. If you're building in the marketplace space, let's chat!

36

7 Comments