Drat, we really thought the Warriors would catch the Pistons tonight! (At least it wasn’t the bloodbath it looked to become at one point.)

Have a great weekend, all.

Top News

Reuters reports that SpaceX pulled in around $8 billion in EBITDA last year on revenue of $15-16 billion, with Starlink accounting for up to 80% of that. It’s numbers like those that have bankers talking about a potential IPO that could raise $50+ billion and value the company at over $1.5 trillion later this year. More here.

OpenAI and Nvidia’s much-hyped plan for Nvidia to invest up to $100 billion and build massive computing capacity for the AI company has stalled, with Nvidia privately signaling the deal was non-binding as investor concerns grow about OpenAI’s trillion-dollar-scale compute commitments. The Wall Street Journal has more here.

Video game stocks slid after Google rolled out an AI model called Project Genie yesterday that can generate playable worlds from prompts, with Take-Two down 10%, Roblox off more than 12%, and Unity falling 21%. Reuters has more here.

Affinity’s 7 Modern Workflows to Win Deals Faster in 2026 

Most private capital firms are sitting on networks worth millions in deal flow. The firms pulling ahead have built systems to actually activate those relationships. In this new guide, Affinity breaks down 7 real workflows used by firms like BlackRock ($13T AUM), Bessemer Venture Partners, SpeedInvest (€1.2B), and Notable Capital to win proprietary deals faster, prevent critical relationships from going cold, and reclaim hundreds of hours per year. The guide goes beyond theory, showing exactly what worked, what didn’t, and why these systems matter heading into 2026. 

A Peek Inside Physical Intelligence, the Startup That’s Building Silicon Valley’s Buzziest Robot Brains

Image Credits: Connie Loizos for TechCrunch

By Connie Loizos

From the street, the only indication I’ve found Physical Intelligence’s headquarters in San Francisco is a pi symbol that’s a slightly different color than the rest of the door. When I walk in, I’m immediately confronted with activity. There’s no reception desk, no gleaming logo in fluorescent lights.

Inside, the space is a giant concrete box made slightly less austere by a haphazard sprawl of long blonde-wood tables. Some are clearly meant for lunch, dotted with Girl Scout cookie boxes, jars of Vegemite (someone here is Australian), and small wire baskets stuffed with one too many condiments. The rest of the tables tell a different story entirely. Many more of them are laden with monitors, spare robotics parts, tangles of black wire, and fully assembled robotic arms in various states of attempting to master the mundane.

During my visit, one arm is folding a pair of black pants, or trying to. It’s not going well. Another is attempting to turn a shirt inside out with the kind of determination that suggests it will eventually succeed, just not today. A third — this one seems to have found its calling — is quickly peeling a zucchini, after which it is supposed to deposit the shavings into a separate container. The shavings are going well, at least.

“Think of it like ChatGPT, but for robots,” Sergey Levine tells me, gesturing toward the motorized ballet unfolding across the room. Levine, an associate professor at UC Berkeley and one of Physical Intelligence’s co-founders, has the amiable, bespectacled demeanor of someone who has spent considerable time explaining complex concepts to people who don’t immediately grasp them. 

What I’m watching, he explains, is the testing phase of a continuous loop: data gets collected on robot stations here and at other locations — warehouses, homes, wherever the team can set up shop — and that data trains general-purpose robotic foundation models. When researchers train a new model, it comes back to stations like these for evaluation. The pants-folder is someone’s experiment. So is the shirt-turner. The zucchini-peeler might be testing whether the model can generalize across different vegetables, learning the fundamental motions of peeling well enough to handle an apple or a potato it’s never encountered.

Massive Fundings

Granola, a three-year-old London startup that records meetings and generates summarized notes, is reportedly in the market to raise at least $100 million at a $1+ billion valuation, or four times the valuation of its May 2025 Series B. Index Ventures is the rumored lead. Forbes has the scoop here.

Oviva, a 12-year-old Berlin company that delivers fully reimbursed digital therapy programs for obesity and chronic conditions through public health systems, raised a $237.1 million Series D round. Investors included Kinnevik, Planet First Partners, A.P. Moller Holding, Lunate, EGS Beteiligungen, and Norrsken VC as well as previous investor Sofina. The round included a mix of equity and secondary shares. Tech Funding News has more here.

Tenpoint Therapeutics, a three-year-old London startup that develops treatments for presbyopia, a natural, age-related decline in the eye's ability to focus on close objects, raised an $85 million Series B round co-led by Janus Henderson, EQT Nexus, Hillhouse, and British Business Bank, with previous investors EQT Life Sciences, Sofinnova Partners, F-Prime, Eight Roads, Qiming Venture Partners USA, AdBio, and Wille also participating. The company also raised $150 million in debt. More here.

Big-But-Not-Crazy-Big Fundings

Poetiq, a one-year-old startup based in Mountain View, CA, that is developing a meta-system to improve large language model reasoning and efficiency across business problems, raised a $45.8 million seed round co-led by FYRFLY Venture Partners and Surface Ventures, with Y Combinator, 468 Capital, Operator Collective, Hico Ventures, and Neuron Venture Partners also investing. More here.

When, a six-year-old Chicago startup that helps employers manage health coverage transitions for workers, raised a $10.2 million Series A round co-led by ManchesterStory and 7wire, with Mairs & Power Venture Capital as well as previous investors B Capital, Enfield Capital Partners, TTV Capital, and Alumni Ventures also taking part. More here.

Smaller Fundings

Arkero, a one-year-old Seattle startup that uses AI to help professional sports teams manage day-to-day business operations such as matchday planning and season-ticket renewals, raised a $6 million pre-seed round led by Game Changers Ventures, with 776, BoxGroup, Garuda Ventures, and Founders’ Co-op also contributing. GeekWire has more here.

Datalinx, a one-year-old New York startup that structures and enriches raw customer data into AI-ready datasets for analytics, personalization, and media measurement across marketing, advertising, and commerce, raised a $4.2 million seed round led by High Alpha and including Databricks Ventures, Berkeley Frontier Fund, and Aperiam. More here.

Sound Games, a two-year-old Seattle startup that develops cross-platform video games sold under a pay-once, play-anywhere model across PC, console, and mobile, raised a $6.5 million seed round led by Point72 Ventures, with Timeless, Daybreak, WOCstar, Hustle Fund, and ZVC also engaging. GeekWire has more here.

Centralized AI-Powered Portfolio Monitoring 

Stop duct-taping emails, spreadsheets, and PDFs. Visible brings all your portfolio data into one centralized, AI-native system. From founder update and board deck extraction to trend detection and MCP server integration, our platform equips 950+ VCs with insights, reporting, and clarity, without chasing metrics. Discover the power of a smarter stack

New Funds

2048 Ventures, a seven-year-old New York VC firm that focuses on leading pre-seed and seed rounds in vertical AI, deep tech, healthcare, and biotech, raised $82 million for its third fund. More here.

Exits

OnlyFans, a nine-year-old London-based creator subscription platform, is reportedly considering selling a nearly 60% stake to Architect Capital, a six-year-old San Francisco-based investment firm, in a deal that would value the business at about $3.5 billion. TechCrunch has more here.

Going Public

Axera Semiconductor, a seven-year-old Shanghai-based company that designs fabless AI inference chips for on-device computing and smart vehicles, is seeking to raise $380 million in a Hong Kong IPO. The company’s backers include Qiming Venture Partners and Tencent. Reuters has more here.

People

Elon Musk discussed plans in private emails to visit Jeffrey Epstein’s Caribbean island in 2012 and 2014, newly released Justice Department records show, contradicting Musk’s public claims that he refused Epstein’s invitations. Business Insider has more here.

On X and in an op-ed in The San Francisco Standard, LinkedIn co-founder and VC Reid Hoffman called on Silicon Valley leaders to stop “bend[ing] the knee” to President Trump, arguing that tech leaders have power, “and sitting on that power is not good for business.” TechCrunch has more here.

Acclaimed director Darren Aronofsky is getting some heat from YouTube trolls for creating an animated series on the American Revolution using AI tools. Variety has more here.

Kofi Ampadu, the partner at a16z who led the firm’s Talent x Opportunity (TxO) fund and program, has left the firm, according to an email he sent to staff that TechCrunch obtained. The move comes months after the firm paused TxO and laid off most of its staff. More here.

Post-Its

Essential Reads

First, the red-hot Clawdbot became Moltbot, then Moltbot became OpenClaw, and now OpenClaw has spawned Moltbook, a social network where AI agents post, comment, and exchange reusable skills through an OpenClaw-powered community layer. Simon Willison’s Weblog has more here.

Apple has lost at least four more AI researchers and a senior Siri executive to rivals including Meta and Google DeepMind in recent weeks, adding to a broader talent drain inside its AI organization amid delays and internal churn. Bloomberg has more here.

Awkward! Instagram will soon allow users to remove themselves from someone’s “Close Friends” list. TechCrunch has more here.

A small Chilean community is launching Quili.AI, an intentionally human-powered chatbot that routes questions to local residents instead of data centers in order to protest the energy and water costs of AI infrastructure concentrated around Santiago. Fast Company has more here.

Detours

Dry January has grown from a niche fitness experiment into a mass cultural ritual, but a widening backlash now frames the practice as performative, joyless, and politically coded.

Brain Rot

Instagram post

Retail Therapy

Bentley just unveiled the 2026 Continental GT S coupe and GTC S convertible, that combine a hybrid 4.0L V8 and electric motor for 670 hp, a 3.3-second 0–60 mph time, and a 190 mph top speed.

Tips (the non-pecuniary kind)

Please send all of your hot gossip to [email protected] or [email protected].

Want to advertise on StrictlyVC?

To book ads directly, contact us at [email protected].

Keep Reading

No posts found