Deal of the Week
Silicon Box raised a $150M round last week. It was co-founded by Berkeley legend Weili Dai, who somehow seems to have a new company featured in this newsletter every few months. Amazing entrepreneur!
Silicon Box is a semiconductor firm focused on chiplets, helping package advanced chips together to deliver cheaper, faster, and more power-efficient systems. The key areas here are the interconnects between the various components and the packaging itself, both of which are becoming increasingly important as AI workloads scale. Everyone is now chasing the next bottleneck in the AI buildout, hence my post last week on the Taiwan and South Korea stock markets. Congratulations to Weili and the Silicon Box team!
Promiscuity of AI Researchers & the Search for Continual Learning
For those who enjoy reading, history, and tech (me!), I would recommend the book Power Law: Venture Capital and the Making of the New Future, which does a whistle-stop tour through the history of Silicon Valley. One of the reasons listed for Silicon Valley’s success versus Boston and NYC is that California does not allow non-competes. In finance in NYC, for example, many high-flying hedge fund traders have aggressive non-competes (don’t feel too bad for these people getting paid a ton for this), but it means it can be hard to port strategies to a new fund as things that work change over time.
I think we are seeing this play out at warp speed in AI. Not a day goes by that I don’t see a top researcher moving from one leading AI lab to another, or Meta paying another few hundred million dollars for a top researcher. This all makes sense, as these top researchers can take their techniques with them, which can make all the difference in the AI race. I just find the whole thing fascinating on many fronts:
Many of these AI researchers are now effectively post-economic, so they want a lab or problem that resonates with them on a personal level.
Seemingly simple things like non-competes can mean industries thrive and push themselves even harder — especially in AI, where the magic, in many instances, is techniques and ways of thinking.
Semi-related to the above, it seems the next big capital infusion into AI research is continual learning. In very simplified terms, this is the way in which a model can learn from new information after being trained and as new information emerges. Many companies have hacks that simulate this, but it is far from fully solved. A couple of startups have recently raised huge amounts of money to go after this, including David Silver. One of the scientists behind AlphaGo, who just raised $1.1B to pursue it.
Another great read is The Infinity Machine: Demis Hassabis, DeepMind, and the Quest for Superintelligence, which charts the history of Google’s DeepMind through the recent history of AI.
Do you really own Anthropic or OpenAI?
Firstly, what is an SPV? It stands for a Special Purpose Vehicle. In simple terms, it is just an entity set up to own a specific asset or investment. SPVs are extremely common across finance and business and can be very useful structures for pooling capital, isolating risk, or making investments more efficient.
However, SPVs are gaining some notoriety because of how they are increasingly being used around hot private AI companies. There is enormous FOMO from investors wanting exposure to names like Anthropic and SpaceX, and a sort of grey market has emerged. Someone gets access to shares through an existing investor or shareholder and then repackages that exposure into an SPV to sell or they start ‘slinging SPVs’ on to other investors.
The problem is that these deals often involve layers of intermediaries. Effectively brokers on brokers on brokers, all taking lots of fees along the way. These are what people refer to as “layered SPVs.” By the time the investment reaches the end buyer, investors may have paid huge fees just for access.
Now there is an even bigger issue emerging. In some cases, investors may not actually own the underlying shares they think they own. Anthropic recently stated that transfers not formally approved by the company could be voided (eek!) and OpenAI also stated this. That means some investors could discover at IPO time that the economics or ownership structure they thought they had is not actually enforceable and their money disappears. It will be fascinating to see how this plays out as companies like Anthropic and SpaceX eventually go public.
To be clear, SPVs themselves are not bad — they are a normal and often valuable financial structure. The problem is paying enormous access fees for exposure to shares that may not even legally transfer. Be careful out there!
GLP-1s Continue to Explode
The explosion of GLP-1’s has got to be looked back as one of the largest changes in the last few decades. This already has ramifications across portion sizes, unhealthy foods. The below is wild showing 1 in 8 Americans use GLP-1s. Many now believes this success will be followed by a boom in peptides with much of it happening on the grey market.
Quick Takes:
Summary by the #️⃣ & 💰:
10 Berkeley-founded companies funded
$285M of capital raised from the 4th April to 10th May
💡 Got any ideas or feedback on how to improve this weekly digest? Just hit reply.
Closed Rounds
🖥️ Silicon Box. $150.6M Early Stage VC 🇸🇬 Semiconductor packaging manufacturer. 💰 Hillhouse Investment Group, Maverick Capital
🐻 Banghua Zhu, Co-Founder & CTO. PhD EECSArticle
🛒 District. $14.7M Seed Round 🇺🇸 Commerce marketplace platform. 💰 Andreessen Horowitz, Greylock, 20VC.
🐻 Edward Koai, Co-Founder & CEO. BA Business Admin Article
🎾 PlayReplay. $12.0M Later Stage VC 🇸🇪 Smart tennis tracking platform. 💰 Alfvén & Didrikson, CSM Ventures, LionTree.
🐻 Hans Lundstam, CEO & Founder. Project Management Article
🛡️ Boost Security. $4.0M Later Stage VC 🇨🇦 Application security platform. 💰 Firebolt Ventures, Hoxton Ventures, White Star Capital.
🐻 Rajiv Sinha, Co-Founder & CRO. Haas MBA Article
🔋 Pollen. $3.8M Early Stage VC 🇵🇹 EV battery-swapping network. 💰 Bynd Venture Capital, KFund
⚡ Celadyne. $0.3M Later Stage VC 🇺🇸 Hydrogen fuel-cell membrane developer. 💰 Breakthrough Energy, Maniv Mobility, Shell Ventures.
🧠 Curiosities AI. Seed 🇺🇸 Voice AI learning companion. 💰 Reach Capital
🐻 Yi Ng, Founder. BS EECS
🐻 Talieh Zargar, Co-Founder & CTO. MS Data Science
🧬 WittGen Biotechnologies. $0.2M Early Stage VC 🇺🇸 AI-powered oncology discovery platform. 💰 Digital Healthcare Partners, Plug and Play Tech Center
🐻 Yun Rhie, Founder & CEO. Haas MBA
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. 🐻
