Paw Andersen

Trying something new this week - a studio industry roundup with my take on what matters. The studio model is no longer an experiment being run in a handful of places. It is infrastructure being built across geographies, sectors, and institutional types. 5️⃣ stories from the past seven days:  1️⃣ OSS Ventures hit a EUR 40M first close on a EUR 75M follow-on fund -- capital raised specifically to back the 30 companies their studio created since 2019. The "build then fund" lifecycle is maturing. B'More Venture Studio with a similar thesis: studio operations first, then dedicated capital to scale what's working.  2️⃣ Delta40 raised $20M from 54 investors across 13 countries to launch Africa's first integrated venture studio and fund. The LP base is almost entirely DFIs and foundations. Impact capital is discovering the studio model as a deployment vehicle, something I seeing across geographies (conversations with Michael Bob Starr and Damias McDonald this week alone) 3️⃣ Start Holdings launched Start.vc out of NYC and Barcelona after a $200M EV infrastructure exit. Another operator-turned-studio-founder. This is the pattern we see quite often in our work at 9point8 -- experienced operators like Jason Goldsmith who've built and exited deciding the studio model is how they want to build next.  4️⃣ UBC HATCH admitted 10 new deep-tech spinouts for Winter 2026 -- mostly climate tech. University venture builders are quietly becoming one of the fastest-growing segments. Follow Evan Allen and Jaimie Testai to keep track of its rapid evolution.  5️⃣ The INVEST Act is advancing through the Senate. If passed, it expands qualifying VC fund size from $10M to $50M and loosens general solicitation rules for accelerator and university events. For studio-funds -- and for folks in the The Venture Studio Forum community tracking policy -- this could be the most significant capital formation legislation since the JOBS Act. If this roundup format is useful, I'll keep it going. What other noteworthy news hit your desk this week, and what stood out to you?

13

1 Comment

Wow! Fascinating article Andrew Antos & Nischal Nadhamuni about how the Klarity platform does knowledge capture! A total disruptor indeed for how consultants traditionally did this work! Capturing not just what happens but why it happens has always been the missing link in enterprise AI, and Klarity ’s approach to surfacing human reasoning through its intelligent platform is a real step forward. The #AI Advisor’s ability to traverse vast context graphs, uncover cross-team patterns, and turn knowledge into actionable insights shows how innovation in knowledge capture can truly transform enterprise decision-making.

8

On-Prem is the New Cloud: Why Enterprises are Reversing Stance in the Age of AI Remember when cloud was the battleground for innovation, agility, and scale? Everyone was tracking the % of on-prem compute moving to the cloud and it was growing every year? Well now there’s a reversal. The reality today? On-prem is the new cloud. Here’s why. AI’s Dark Side: Legal Risk and Data Governance Headaches Enterprises are excited about AI, but also uneasy as vendors like OpenAI face intensifying legal pressure. In a high-stakes lawsuit, a judge ordered OpenAI to preserve all ChatGPT chats (even deleted ones), overriding its 30-day deletion policy. Imagine that: you delete sensitive conversations, but they’re still sucked into legal limbo. The precedent is clear: AI prompts and outputs are discoverable records, demanding integration into enterprise ESI (Electronic Stored Information) policies. For large organizations handling sensitive data, the risk is simple: you can’t afford your AI tools to turn into legal liabilities. Especially when global regulations (GDPR, etc.) and internal governance collide with unpredictably broad discovery orders. Sovereignty, Control, Trust Many enterprises still ban ChatGPT or impose strict rules on usage. Even when allowed, token limits are throttled. Enterprises are cautiously letting AI in (as they should). Enterprises need to have control and trust. SAP is leaning into this with its Sovereign Cloud On-Site, deployable on customer premises with SAP-managed infrastructure. It is built to deliver data, operational, technical, and legal sovereignty in one package. Doubt it will be the last. Scaling Pains Talking to founders, three reasons stand out why cloud isn’t scaling with AI: 1/ The rate at which AWS/GCP lets you provision compute and the compute it wants you to provision both falter as AI and AI driven developers deploy a lot more code more frequently than before. So sometimes it’s not about data governance, but actually about performance and functionality. 2/ Shadow AI is rampant—employees paste sensitive info into models despite restrictions—creating demand for better permissioning tools. 3/ Data leakage remains a live risk, like the August 2025 Grok incident where hundreds of thousands of private chats became publicly accessible and indexed. One thing is for sure - founders selling into the enterprise are scaling faster if they offer the ability to run on-prem. The Future is Hybrid Don’t get me wrong - cloud is not going anywhere. But neither is on prem. A few years agoI thought we were headed for a full cloud shift, but now I think the future is truly hybrid.

34

4 Comments

Key takeaways: - The core shift is economic, not hype. LLMs turn “AI” into a recurring operating cost, so pilots quickly become capacity planning, governance, and ownership. - Power and timelines show up because production inference needs sustained compute capacity. When usage increases across the org, you run into procurement, security review, integration work, and infrastructure readiness challenges. - The winners are not the teams with the flashiest demos. The winners are the teams that convert compute into measurable outcomes inside a workflow with traceability and controls. - In 2026, “AI strategy” that is not tied to a specific workflow, KPI, and operating model is mostly theater. If you are building the application layer on top of the datacenter buildout, execution systems, workflow automation, data-layer tooling, compliance automation, I want to see what you are shipping. Founder intro: www.axion.ventures

6

a friend just texted me: "Hearing there are 'senior devs' who are very skeptical of AI generated code (aka not using any AI tools) at a startup I’m interviewing with… interesting that they are still employed in a startup against this backdrop [see image]" my response: seems like there are just different games being played those who embrace AI as a medium for expressing ideas are playing the right game — they're not just treating it as a "better tool" if you've been in industry awhile, you may need an intellectual reboot to see it last night i went to a VentureBeat event; mostly enterprise folks… several people (from IBM, Walgreens, etc) asked about “change mgmt” and how to convince their teammates to adopt AI tools… someone raised the question of brown-field vs green-field opportunity w/r/t AI… is it worth it to upskill existing workers or better to just start fresh…? Great question! for people entrenched in a job fiefdom, e.g. where the power players and how to execute reliably, the perceived risks of novelty are infinite. they're not wrong, they're just not interested in changing the game they've mastered. aaand also, developers are the self-disclosed laziest people in the world. but they're incredibly productive because they'll happily spend a week automating some repetitive task so they never have think about it again. this is why generative AI and coding go so well together: the existing developer culture despises inefficiency, and is always looking for ways to cut corners in favor of a more elegant, efficient route. this is because developers are incredibly sensitive to the tiniest impediment to achieving persistent flow states. it's like noticing a typo in that green-on-black matrix screensaver and instead of ignoring it as a minor irritant, they'd prefer to rewrite the screensaver from scratch to fix it. but for non-developery roles, this gameplay is totally foreign. maintaining those impediments ensures job security. since they know how to navigate around known obstacles, their tacit knowledge is an insurance policy against obsolescence — for now. so it makes that sense that some senior devs, whose sunk costs could represent hundreds of thousands or millions of $$$ into existing, _stable_ systems might resist the quasi-non-deterministic output of AI coding tools. except (as they say) there is more than one way to skin a cat. here's a new game to invite resistants to play: have the skeptical senior devs review several blind PRs, all submitted by “new,” unfamiliar contributors (a mix of humans and AIs), and see if they prefer the human or AI-generated code. if the AI code really is as good as many of us believe, the experience might snap them out of their sunk-cost fallacies. because until you surprise yourself with what these tools can actually accomplish—in a blind taste test—you’ll never really see what winning this new game looks like, even if it's happening right in front of your eyes.

21

14 Comments

Worth paying attention to... Anthropic vs. Cursor. While Cursor was paying Anthropic hundreds of millions in API fees to power its AI code editor, Anthropic seems to have studied exactly what developers wanted, then launched Claude Code to "eat" their biggest distribution partner. The result. Claude Code hit $1B ARR in just six months and has already surged past $2.5B, proving that when you're a "wrapper" startup, you're often just paying your provider to learn how to replace you. Keep an eye on Cursor layoffs...

89

13 Comments

As 2025 drew to a close and most people were heading out for the holidays, Sam Altman, and I were discussing how we could unite the world’s leading AI models with the world’s fastest processor and unleash a new era of creation and discovery.   Ideas like that always seem impossible… until done.    We did it.   Today, OpenAI and Cerebras announced they have signed a multi-year agreement to deploy 750 megawatts of Cerebras wafer-scale systems to serve OpenAI customers. This deployment will roll out in multiple stages beginning in 2026, making it the largest high-speed AI inference deployment in the world.   I knew Andrew Feldman was different the moment I first met him in 2016. Big ideas like this come to life when people think (and build) different.   It wasn’t just that he was insanely smart. It was his unshakeable conviction in his “crazy” idea. It was his intense belief that it was actually crazier to do nothing than to try to bring his idea to life, risking everything he had in the process.    The same is true of Sam Altman. Like Andrew’s vision for AI hardware, Sam’s vision for AI has always been audaciously bold, big, and difficult. Launched around the same time, OpenAI and Cerebras have been on parallel missions to make AI faster, smarter, and more accessible. They’ve never given up, even when it has truly seemed impossible.     I’m drawn to people who manically pursue their mission. It’s the way I and the Eclipse team operate too. It’s the only way to take the big swings that fundamentally transform industries — and mindsets.  That’s why, to me, it seemed natural for Andrew and Sam to work directly together to take AI to the next level. They’ve long supported and collaborated with one another, but integrating their technology was always the next frontier.    When the three of us began talking about what this partnership could look like, we felt that familiar buzz that has driven each of us our entire careers: We knew this was a radically ambitious idea. Some might even call it crazy. That’s how we knew it was exactly the type of idea worth going for.    Andrew and Sam are trailblazers who defied the limits of hardware and software to get AI where it is today. Together, they can shape how AI transforms our lives, our country, and the world of the future. I’ve never been more excited to partner with our portfolio companies to rebuild and redefine the industries that matter. More from The Wall Street Journal's Kate Clark and Berber J.: https://bit.ly/3NL0ETt

554

22 Comments

Voice is a super interesting modality right now - maybe the first modality we're seeing move to open source models across a number of scale ups / enterprises. Reliability concerns, high costs, and open source model performance are pushing engineers to do their own fine tuning vs. relying on third-party vendors of proprietary models. Many of these orgs have already been collecting their own first-party data and now with third-party vendors like Extrian, David AI, etc they can train really high quality models. RL has been insanely hyped, but it's been unclear how long it will take scale ups and enterprises to actually lean in. Voice AI might be hitting that inflection point faster than expected.

45

7 Comments

If 2024 was the year that I called BS on corporate AI theater, 2025 was the year I yawned and searched for substance elsewhere. * Voice IVA was overfunded, overhyped and did not automate away the industry like the YC kids wanted us to believe. * “Agentic Orchestration” is still an abstract concept and demo-ware at best. * Agent Assist is useful, but only for a small percentage of interactions that require depth or non-routine complexity. * Automated QA has become a checkbox illusion of AI adoption, where the noise-to-value ratio is inverted. * Post-call summaries matter little without intentional construction and automated data pushes into the CRM. * Organizations still lack the know-how to implement and adopt AI with any measure of success, while most vendors continue pandering to the press. So, in 2026, I'm putting the "Last Mile, First" by attacking the constraints to customer adoption. Our thesis is Unified CX. Customer experience breaks when strategy, tooling, deployment, and operations are split across vendors with misaligned incentives. * Consultants advise without ownership.  * Integrators deploy and then leave.  * BPOs run agents but seldom own the full stack. * Software vendors collect license fees whether the thing works or not. Nobody owns outcomes. So nobody delivers them. That's why I'm converging our separate AI-CX companies into one: InflectionCX One system. One partner. End-to-end accountability. We discover opportunity, design the solution, deploy the technology, and operate it in production with live Agents and AI ones -> converging people and AI into a single operating paradigm. And, we’re running the opposite playbook of everyone else: * We’re investing heavily in non-CX-facing systems that expand internal capacity and enable precision and scale: recruiting, learning management, and go-to-market. * We’re putting Foundation over foundational models, ensuring that your data, context, governance, and processes exist before AI is deployed. If those aren't ready, neither is your AI. * Decades of digital red-lining have left large parts of the country without residential high-speed internet. No internet means no remote work. So we’re putting call centers where they matter most: Rural America. Which also means that next year, I get to go play in the dirt. 

54

8 Comments