Dec 5, 2025 · 53m · neon-show
From -1 to 0: How Founders Navigate the Earliest Stage with Aditya Agarwal & Prateek Mehta | SPC
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of The Neon Show, South Park Commons partners Aditya Agarwal and Prateek Mehta discuss navigating the pre-idea 'minus one to zero' stage of entrepreneurship, how AI is compressing startup lifecycles, and the massive untapped potential of India's frontier tech builders.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
Siddhartha as informed peer 4.2 Guest teaching 5.0 Guest disagreement 1.2 Siddhartha pushing back 1.3
speaking balance: gold is Siddhartha, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 43:25 Aditya rejects the pessimistic India AI narrative
Aditya directly attacks prevailing skepticism, calling the idea that India cannot build foundational AI 'ridiculous' and stating founders need no permission.
Hardest push from Siddhartha ▶ 41:57 Host challenges guests on Indian talent parity with Silicon ValleySiddharth directly presses the guests to be honest about whether Indian tech talent genuinely matches the best builders in the Bay Area.
Biggest teaching moment ▶ 11:55 Prateek reframes the definition of community successPrateek corrects the conventional expectation that everyone must launch a venture, clarifying that achieving conviction—even deciding to take time off—is a top-tier success metric.
Siddhartha holds their own ▶ 26:09 Host demonstrates VC expertise with proprietary founder insightSiddharth cites a specific deal won by Neon Fund against Tier 1 funds because experienced founders reject dogmatic VC preaching in the AI era.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Siddhartha as informed peer | Guest teaching | Guest disagreement | Siddhartha pushing back | Why |
|---|---|---|---|---|---|---|
| Episode Preview and Highlight Clips | 3 | 6 | 1 | 1 | Siddharth introduces the episode and asks Aditya to define the 'minus one' stage. Aditya gives an extensive narrative on his background at Facebook and Dropbox, explaining how SPC was born as an exploratory learning club. | |
| Evolving from Learning Community to Seed Fund | 4 | 6 | 1 | 2 | Siddharth asks how capital fits into what sounds like an altruistic builder community. Prateek and Aditya explain how backing founders at the exploration stage led organically to creating pre-seed and seed funding rounds. | |
| High Talent Density and Conviction-Driven Outcomes | 3 | 5 | 1 | 1 | Siddharth asks about conversion rates from joining to starting companies. Prateek details the strict sub-1% acceptance rate and explains that reaching clear conviction—whether starting up or recovering—is the true metric of success. | |
| Scaling SPC Bangalore and Running VCs Like Startups | 3 | 5 | 1 | 1 | Aditya and Prateek discuss rapid growth in Bangalore from an expected 20 members to 70-80, emphasizing that VC firms themselves should operate like agile startups rather than rigid institutions. | |
| How AI Shifts Startup Velocity and Opportunity Frontiers | 5 | 5 | 1 | 1 | Siddharth asks how AI changes early-stage investing and product development. Aditya elaborates on rapid iteration cycles and how platform shifts compress the journey from seed to Series B. | |
| AI-Forward Operations and Leading with Curiosity | 6 | 4 | 1 | 1 | Siddharth shares an anecdote from Neon Fund about founders rejecting dogmatic Tier 1 VCs. Aditya and Prateek enthusiastically agree, detailing their unpretentious office philosophy and internal AI tooling. | |
| Successful Portfolio Companies Born at South Park Commons | 3 | 5 | 0 | 0 | Siddharth asks about top portfolio outcomes from SPC's history, prompting Aditya to list Base10, Render, Pilot, Gamma, Luma AI, and Profound. | |
| Expanding South Park Commons to India | 4 | 5 | 2 | 2 | Siddharth questions the timing of expanding to India. Aditya and Prateek explain the catalyst of Binny Bansal's operational encouragement and the rising cohort of repeat founders looking to build globally. | |
| Debunking Doubts Around India's Deep Tech and AI Talent | 7 | 5 | 3 | 3 | Siddharth probes whether Indian talent truly matches Bay Area builders and shares data points of tier-2 AI adoption. Aditya forcefully rejects the narrative that India cannot build world-class AI, citing deep mathematical foundations among local engineers. | |
| Physical AI Frontiers and Concluding Thoughts | 4 | 4 | 1 | 1 | Siddharth asks about future value creation frontiers like physical AI and agents. Aditya and Prateek discuss robotics, science fiction coming to life, and long-term optimism for Indian tech value creation. |
Statements from this episode (13)
Agarwal: Minus-one learning must be paired with weekly building and demos
“For us, you learn so you can build. So every week you have to build something with velocity, with speed to demo to everyone.”
Agarwal: Vitalik Buterin and Jennifer Doudna spoke to early SPC learning club
“So, you know, in the early days, someone like you know, like a Vitalik Buterin who was just building the ether network, who obviously became Ethereum. Someone like a Jennifer Doudna who was doing groundbreaking work on CRISPR. They would come to a small kind o…”
Mehta: SPC invests $400K at pre-seed and $600K at seed
“And typically what we do is we commit a million dollars to a team at the pre-seed stage. So, 400 K at the pre-seed stage and at least 600 K at the seed.”
Agarwal: Start a company only if you accept 20 years of pain
“Not everybody should start a company, right? And I think that our environment, when somebody comes in to join the community is that you should start a company. If you're willing to go through 20 years of pain, Right. But sometimes honestly, it's like better to…”
Mehta: South Park Commons' selection rate is under one percent
“The selection rate for the community is less than one percent, right? So by way of number of applications that come in and the number of people who actually get an access to the community, right?”
Agarwal: SPC Bangalore added 70 to 80 community members in its first year
“In the last one year, we've been very, very, it's been amazing to see, like we've had roughly about 70 community members. 70 to 80 community members join us in SPC Bangalore.”
Agarwal: 15 to 20 companies emerged from SPC Bangalore in year one
“And we've had roughly 15, 20 companies come out of it, which we are very, very excited to be partnering with.”
Agarwal: Organizations run on small implicit signals, not high-level stated values
“I believe a lot that organizations run on all the small signals that you send, not kind of like the high level values. And then the values are important. Values and principles are very important, but it is almost like the implicit signals that the environment …”
Mehta: SPC rejects fixed market theses for "axes of curiosity"
“When we invest, we don't, we're not talking about thesis and right. This market is going to go this way. We're talking about access of curiosity, which is something that we publish every six months. We talk about what are the broad areas where we believe thing…”
Agarwal: SPC has produced four to five foundation model companies
“And then we also have had probably like four foundation model companies come out of like essentially SBC four or five actually kind of doing everything from general purpose kind of like open source models to more specific models around essentially everything f…”
Mehta: Failures have anecdotal value, but success teaches like nothing else
“Failures have high anecdotal value, but success teaches you like no one else.”
Agarwal: It is "ridiculous" to think India lacks talent for AI
“I strongly disagree with the narrative that a bunch of people have that India can't do AI. I think that's ridiculous. I think we have seen enough examples from some of our members, some of the companies that they are starting. That India actually has absolutel…”
Agarwal: Average American CS grads lack the advanced math for AI models
“Honestly, you ask the average CS graduate in America to understand, like, you know, matrix, you know, kind of, like, matrix math, advanced linear algebra, kind of, like, to understand, kind of, like, how these actually models work, they would struggle with it.…”