Humans and AI have something random in common
Necessary Nuggets
Happy Wednesday! If you’re new here, welcome to Necessary Nuggets, your one-stop pre-seed shop. We deliver updates from Necessary Ventures and helpful tidbits on our little corner of the world. Every edition is also on our blog.
Jobortunities 🚨
✨NEW FEATURED ROLE✨
Nevoya: First fully-electric trucking fleet in the United States, helping truckers trade gas-guzzling for environment-protecting. Get involved and tackle unsolved problems in the trucking industry.
Necessary Ventures: If you love this newsletter, we think you might like investing in what the world needs too. We have full-time and part-time opportunities to think early stage and have an impact on solving founders’ problems.
| Investor |
Taro: Taro increases access to healthcare while reducing costs and physician burnout via direct primary care based health insurance.
| Product Engineer |
Forage: Create a unified API that helps low-income Americans buy online groceries. It’s Stripe, but for Electronic Benefits Transfer.
Reach out with any questions! All 164 open job postings are listed here.
Market Stirrings 🚩
Here's what the week looked like in pre-seed:
$3.5M
Total Amount Raised
15
Total Funding Rounds
$236k
Average Dollars per Round
$1.2M-$2.4M
Estimated Valuation Range
Data aggregated from proprietary research and Crunchbase; valuation estimate based on 10-20% ownership stake.
Healthy dilution
Early-stage health startups are seeing less and less dilution compared to SaaS startups, experiencing 5.8% less dilution as of Q1 2024. This is a relatively recent trend, with health startups starting to experience less dilution beginning in Q2 2023. The trend doesn’t hold for later stages, with Series A and B data showing higher dilution compared to SaaS.
Good Reads 📖
For the rushed reader …
Fulcrum BioEnergy edges towards demise after failing to find a cost-effective model to make good on its promise of turning garbage into fuel.
Engineers at Gramener, a data analytics and engineering company, discovered that LLMs have favorite numbers, just like people do.
Clean Energy Ventures (CEV) is stepping up its game and just closed a $305 million second fund, jumping up substantially from its debut fund at $110 million.
For the less rushed reader …
LOSING ENERGY: What happens to a startup deferred? Does it dry up, like a raisin in the sun? Or fester like a sore, and run? Both might be true of collapsing climatetech startup Fulcrum BioEnergy. Fulcrum BioEnergy, which previously raised $1 billion from industry giants like BP and United Airlines, now edges towards demise after failing to find a cost-effective model to make good on its promise of turning garbage into fuel. The startup tripped over a series of hurdles along the way. Fulcrum originally estimated plant operations would begin in 2010, but that was delayed four times, spanning 12 years total and actually starting operations in 2022. Even then, the picture was less than rosy -- soon after the plant opened, Fulcrum realized that it was producing nitric acid, which ate through its fancy new equipment and blasted harmful nitrogen oxide emissions into the air. When it came time to pump out the fuel, it could only send out 350 gallons -- the size of a small hot tub in the middle of an almost empty railcar. Fulcrum laid off most of its employees (standing at about 100 total) and defaulted on $289 million in bonds. A sustainable airline industry might not be taking off any time soon, especially with SAFs already in short supply.
RANDOM THOUGHTS: If you're looking for a random number, stop talking to ChatGPT and watch the Mega Ball instead. Engineers at Gramener, a data analytics and engineering company, discovered that LLMs have favorite numbers, just like people do. Conducted on several major chatbots, including OpenAI's GPT-3.5 Turbo, Anthropic's Claude 3 Haiku, Google's Gemini 1.0 Pro, the experiment showed that when asked to pick a random number from 0 to 100, all three models picked one number more than any other. Moreover, there was a set of about 10 numbers chosen with higher frequencies. In case you're wondering for your next gamble on the Mega Ball, GPT-3.5 really likes 47. The bias demonstrated by every bot was similar to humans, clearly avoiding the low and high numbers like how humans would avoid picking 1 or 100 when asked the same question. The study brings up questions of how we can train AI in a way that prevents pseudanthropy, an unnecessarily long word meaning the impersonation of humans. One thing seems certain: LLMs are proving themselves to be masters of deception, and perhaps there's some Harry Houdini hidden in their code.
CLEANING THE FLOOR: A new venture fund is raising the roof with fresh capital. Clean Energy Ventures (CEV) is stepping up its game and just closed a $305 million second fund, jumping up substantially from its debut fund at $110 million. The firm targeted strategic LPs for its newest fund, leaning into a consortium fundraising model where the firm's LPs also represent some of the largest clients of its portfolio companies. That model is similar to that used by Fifth Wall and may be advantageous for facilitating distribution deals and information sharing. In climate tech especially, targeting corporate strategic partners across industry verticals can be key for startups when it comes time to fundraise. Early-stage climatetech startups looking to get a piece of CEV's pie must agree to relatively stringent requirements: portfolio companies must commit to reducing greenhouse gas emissions by a whopping 2.5 gigatons over 25 years. For those looking to compliment their CV with an investment from CEV, it takes faith, trust, and gigatons of hard work.
Fire Up the Pre-Seeds🔥
Notable pre-seed raises in …
Climatetech
Syntopa - Massive goals for microbes
The Problem: Agriculture accounts for about one-quarter of the world's greenhouse gas emissions.
The Tech: Microbial products that enhance rock weathering for carbon capture.
Recently Raised: $575k in funding led by Ponderosa Ventures.
Also Note: After receiving his PhD in plant genetics from Berkeley, founder Hans Holtan worked for two biotech startups, learning how synthetic biology can be applied to make microbes better.
EcoRatings - Data makes the world go green
The Problem: Organizations are setting aggressive sustainability goals and require more advanced ways to measure their progress.
The Tech: A generative AI platform that uses machine learning and big data to quantify the environmental impact of products and services.
Recently Raised: $1 million in funding with contributions from We Founder Circle, 888 VC, Vinners, Indigram Labs Foundation, and Google.
Also Note: The startup claims to already have over 50 paying clients.
Outro🚪
Feel free to reply to this email with all questions, feedback, or comments. We’ll be iterating and curating the NVTC newsletter according to your interests.
Some last matters of business:
If you’re a technologist (engineer or product manager / designer with a technical background) join us on the NVTC LinkedIn group if you haven’t. We’d love to have you!
Sign up here if you’re interested in co-investing with Necessary.
If you’re a startup founder, we’d love to help where we can! Brex provides full-stack finance solutions for startups. Sign up via Necessary and get bonus points.
Thanks for reading, and see you next week!
© 2024 Necessary Ventures
San Francisco, California