Everything but SaaS: Where Tech’s Next Big Bets Lie
AI, Web3, and Climate Take Center Stage - as a Trinity
Last week, I attended a panel discussion hosted by South Park Commons (SPC) and Union Square Ventures (USV) — early investor in Twitter, Zynga, Coinbase, Etsy, Tumblr, etc. as well as Slope, where I work.
Somewhat defiantly titled “Everything but SaaS”, the talk was about where tech is headed outside the usual SaaS arena. The speakers discussed three themes that USV is focusing its bets on — AI, Web3, and climate — and how they closely relate to one another.
I admit - prior to this, I had never associated Web3 with either AI or climate tech. I was intrigued by blockchain, but thought most Web3 startups were either BS, or riding the Bitcoin bubble, or the latest versions of financial arbitrage at its worst (yes, I meant you, DeFi…).
This talk, however, gave me a brand new lens through which to look at Web3. Plus plenty of AI insights - one of my favorite topics.
Below, I’ll share my key takeaways.
Disclaimer: while I try my best to be accurate, my interpretations do not necessarily equal speaker views.
Thank you — Rebecca Kaden (Partner, USV), Nick Grossman (Partner, USV), Andy Weissman (Managing Partner, USV), and Arian Agrawal (Co-Founder, SPC) — for the ideas-packed 45 minutes on a freezing Tuesday night.
SaaS Will No Longer Be the Same
SaaS opportunities are not gone, but getting “edgier.”
AI could massively popularize voice-based human-software interfaces, for both consumers and businesses. We may be heading to a prompt-first world (versus command-first).
Building new apps with AI is so easy that one could spin up a no-code product over a weekend. Future apps will be more like “content” than full platforms.
USV’s Thematic Framework
USV filters investment themes on three criteria:
Large market
Edge capabilities
Changing industries or user behavior under pressure
Their conclusion? AI, Web3, and climate — three complementary themes, closely interwoven in the same fabric of data and intelligence.
AI <> Web3
AI generates infinite content; Web3 can add “ground truth,” scarcity, and “hardness”.
Web3 is inherently “machine-only”: permissionless, inter-operable, and open—ideal for AI agents.
Web3 can be hard for humans to grasp, but straightforward for AI. This could mean natural synergy, similar to LLM applications in law (think Harvey).
E.g., AI agents can pay each other seamlessly in crypto, without the normal frictions of traditional, human-based payments.
Crypto networks might enable decentralized AI compute (like model training), which could limit Big Tech’s monopoly.
Climate <> Web3
Crypto protocols can finance green energy infrastructure in ways tradFi struggles with.
New incentive frameworks could lead to bigger climate impact.
Climate <> AI
Climate tech can supply green energy for AI compute, which is notoriously power-hungry.
All Three Together
Each domain—AI, Web3, climate—can amplify the others.
E.g., by routing energy more flexibly, Web3 could help balance grid load when AI-driven power demands spike.
Audience Q&A Highlights
Policy and Regulation: Future for These Policy-Sensitive Sectors?
Web3 regulations ebb and flow. We need clearer rules so more builders can feel comfortable entering the space.
AI regulation is a wildcard—nobody’s sure how to approach it.
Historically, regulations lagged behind new tech. In fact, new tech forces regulatory evolutions. E.g., companies like Airbnb and Uber helped shape new laws — sometimes after breaking old ones.
Not that breaking laws is good; however, there’s so much legal ambiguity for disruptive technologies, that startups will never take off if they cannot stomach any such risk.
USV will take on regulatory risk with founders who have their “eyes wide open and are prepared to fight.”
AI “Gold Rush”: Where Are the “Shovels”?
In other words, which opportunities are attractive because they enable the proliferating AI builders?
Hard to tell which AI tooling and infrastructure will be absorbed over time by major model providers like OpenAI. The faster foundational models improve, the greater such threats are.
USV hedges by investing in areas less likely subsumed (e.g., energy for AI, niche foundational models).
Crypto-based AI infrastructure could be a defensive opportunity, given it’s based on technologies completely different from GenAI.
Verticalized AI applications may also be more protected.
Evaluating Investments
USV looks for ideas that are “native” to a brand new tech paradigm.
“Native” means something that was (and likely still is) non-obvious, but will be a no-brainer tomorrow, when the emerging paradigm becomes the norm.
E.g., think Instagram and endless scrolling on a mobile screen.
Focus on how people might use AI in ways they never did before, because AI simplified something and made it 100x more accessible. The real demand could be much bigger than we expect.
Precedents: Airbnb enabled countless casual landlords, and Substack democratized paid writing.
AI Threat to SaaS Pricing Power?
Technology is inherently deflationary.
That said, it shrinks some value pools but expands others. USV keeps track of how AI might shift where value accumulates.
Intersection Between Hardware, Software, and Services
Interesting synergies and crossover opportunities.
Software is making hardware cheaper. Hardware can anchor embedded SaaS (e.g., Toast’s POS + SaaS for restaurants). (Me: in turn, software can also differentiate hardware - think Nvidia’s GPUs and CUDA)
Creative business models emerge that combine software and services (e.g., USV has backed Teamshare, a company rolling up small businesses and providing software, business services, and financial products).
Behavioral Change for AI Adoption
Creativity and gamification in product design can help drive desirable user behavior changes.
No hardware “entry barriers” for AI (e.g., everyone already has a smartphone) means adoption can be rapid.
Enterprises move much slower. USV doesn’t do much enterprise SaaS because that can be a bottleneck.
Removing middlemen is often the opportunity for AI (enterprises can act as gatekeepers).
Overcoming Intractable Complexity
AI could greatly simplify tasks that otherwise are complicated for human - deciphering healthcare insurance, implementing ERP integrations, etc.
Interoperability of data across app silos—once a big hurdle for both consumers and businesses—might be solved by AI. (E.g., Glean for enterprise search)
Overall, USV believes AI, Web3, and climate form a new frontier. The panelists expect these fields to reshape how we generate content, verify it, power it, and finance it—all at once.