Deal of the Week
Databricks is charging through the alphabet, closing a $3 billion Series M at a monster $188 billion valuation. The private markets just keep getting bigger and bigger. You can read an earlier post on what Databricks does here.
One of the most interesting details is that CEO Ali Ghodsi reportedly said the fundraise was driven by customer demand and the need to buy more GPUs. Jensen is likely licking his lips.
While parts of the AI infrastructure buildout may be experiencing a few wobbles, see the Kimi post below. The broader investment cycle appears set to continue. Databricks feels like one of the major beneficiaries, providing the data and AI platform enterprises need to build and deploy their AI applications. The other big winner seems to be Clickhouse.
Congratulations to the team on the fantastic continued progress.
Chinese Models to be Banned?
The big news last week was that Moonshot AI, a Chinese AI lab, launched an open-source model Kimi K3 that appears to be at the cutting edge of AI, based on the leaderboard from Berkeley-founded Arena.
It was so popular and Moonshot AI so compute constrained that they had to quickly limit access. We need more chips!?
This is a serious shot across the bows of the US AI labs. Kimi claims to offer frontier-level performance at a much lower cost. Although the cost advantage looks less dramatic when adjusted for intelligence per token. More importantly, it challenges the idea that the US frontier labs are consistently three to six months ahead.
My suspicion (although I would not support it) is that we will eventually see some form of restriction on Chinese open-source models. This might even be from the Chinese themselves!
One gauge I have been looking at is Alibaba’s share price. Alibaba owns around 36% of Kimi’s parent company, Moonshot AI, and 17% of MiniMax, another leading Chinese AI company. It has also announced that its next Qwen model will be on par with nearly all existing frontier models.
See what they own below.
This comes on top of a reacceleration in Alibaba’s cloud revenues. Yet despite all of this, the stock is down 22% YTD. That does not look like the sort of AI fever you would expect for a company with exposure to several of the world’s leading AI models.
I believe the market may be pricing in some form of future restriction, particularly as Chinese labs have been accused of distilling US models. Distillation is essentially where you send hundreds of thousands of prompts to an existing AI model and then use the outputs to help train your own. This is one explanation for how Chinese labs may have reached the frontier at a much lower cost.
The deglobalistion of data is set to continue—even as the price of intelligence keeps getting cheaper.
The AI race to where?
To be clear, I am team USA in this race.
But something on my mind is where is this race to?
So much is written about the great AI race between the US and China. Every era in history seems to manufacture some form of competition between the great powers, so I thought it would be interesting to look at a few historical precedents and see how they ended.
Sadly, many of these great races seem to end in one of three ways: war, economic exhaustion or an uneasy equilibrium.
A few examples below:
But if the AI race is really a race to dominate intelligence, I am a little perplexed by what victory would actually look like.
The most optimistic scenario may be that one technology ecosystem becomes dominant, in the same way the US-led internet and software ecosystem did. But as the post above shows, that outcome currently looks far from certain with progressively better models from both sides.
And unlike the Space Race, there is no obvious finish line. There is no single Moon landing moment after which everyone agrees the race is over. Intelligence is continuous, capabilities diffuse and any lead may only be temporary.
So the question that really needs to be answered is how does this deescalate if no side believes it can afford to stop.
Quick Takes
Great article on why unit economics and margin structures may be coming for technology.
Pretty impressive from Haas Grad (podcast here) Jeff Totten who started Evergreen (a private equity roll up) that now does $1.5B in sales and $250M in EBITDA. They bought 47 companies last year!
Great to see Berkeley company GC AI listed as a legal startup to watch! I am grateful to be a small investor.
The rumour doing the rounds is that Anthropic will buy Berkeley founded Physical Intelligence (who recently raised $600M). They are building the foundation model for robotics. 2026 is really shaping up for Robotics as we suggested in our 2025 round up where we flagged Physical Intelligence. ! (Checkout older posts on Robotics here)
Summary by the #️⃣ & 💰:
10 Berkeley-founded companies funded
$3.4B of capital raised from the 13th July to 19th July
💡 Got any ideas or feedback on how to improve this weekly digest? Just hit reply.
Acquistion
💊 Swing Therapeutics. Acquisition 🇺🇸 Digital treatments for chronic conditions. 💰 XRHealth,
🐻 Mike Rosenbluth, CEO & Founder. PhD Bioengineering. Article
Closed Rounds
🧱 Databricks. $3.0B Series M 🇺🇸 Unified data and AI platform. 💰 Coatue Management, Andreessen Horowitz, Dragoneer Investment Group.
🐻 Ali Ghodsi, CEO & Co-Founder. Visiting Scholar & Professor. Article
☢️ AdvanCell. $315.0M Series D 🇺🇸 Targeted radiopharmaceutical cancer therapies. 💰 Ally Bridge Group, Alpha Wave Global, Bain Capital Life Sciences.
🐻 Andrew Adamovich, Founder. BA History. Article
🧩 TYLsemi. $43.0M 🇺🇸 Chiplets for custom AI infrastructure. 💰 Matter Venture Partners, Viola Ventures, Egis Technology.
🐻 Sunil Bhardwaj, Co-Founder & COO. Haas MBA. Article
📡 MyDecisive. $12.0M Early-Stage Round 🇺🇸 Autonomous AI telemetry control hub. 💰 Copper Sky Capital
🧬 Pluristyx. Series A 🇺🇸 Stem-cell therapy development platform. 💰 BioLife Solutions, BroadOak Capital Partners.
🐻 Jason Carstens, CEO. PhD Chemical Engineering. Article
🧭 Silicon Microgravity. $8.0M 🇬🇧 GPS-free precision sensing technology. 💰 West Hill Capital, UKI2S, Future Planet Capital.
🐻 Ashwin Seshia, Co-Founder & CSO. PhD & MS EECS. Article
🎓 Section. $3.2M 🇺🇸 Enterprise AI adoption platform. 💰 Activant Capital, General Catalyst, GSV Ventures.
🐻 Scott Galloway, Founder & Chairman. Haas MBA. Article
🦠 Affinity Biosensors. $2.9M 🇺🇸 Rapid antibiotic-response testing devices. 💰 Founders Capital Partners, Rose Ventures
🐻 Ken Babcock, Founder & CEO. AB Physics
🤝 Mio. $2.2M Pre-Seed 🇫🇷 Slack-native AI workplace colleague. 💰 Fabric Ventures, Topology Ventures
🐻 Arthaud Mesnard, Founder & CEO. BA EECS. Article
💡 Ingantec. $0.2M Seed Round 🇺🇸 High-efficiency red microLED materials. 💰 Lubar & Co.,
🐻 Alan Yeung, Co-Founder & CEO. Haas MBA.
Date Built By Berkeley Started | Companies Funded | Total Raised ($M) |
7/8/24 | 808 | 230,792 |
Our goal is to document the startup ecosystem of Berkeley-founded companies. Please share this newsletter with any Cal Bears in your network so we can crowdsource information about all investment rounds and job opportunities.
Did we miss a company or want to announce a round? Add it here, and we will post it next week.
Do you have a job you want to post from a Berkeley Company? Add here.
Is there someone in the Berkeley ecosystem that you would want us to do a profile on? A founder, funder, or general startup person? Add here.
Built By Berkeley is not affiliated with UC Berkeley, but maybe we will be one day if we get enough subscribers….
Built By Berkeley, where we announce all the funding rounds by Berkeley-founded companies. This is a community effort, so please let us know if we missed a company here. 🐻
