Most startups would be better off without venture capital.

Lance Cottrell, founder of Feel the Boot, Anonymizer founder, and president of North Bay Angels, built an internet privacy company that exited in 2008.

On the Ignite Podcast, Lance joins Brian Bell to challenge conventional startup wisdom.

Subscribe now

We discuss:

• Why Anonymizer raised only $2.5 million, exited with millions in cash, and never needed another funding round after finding product-market fit.

• Why a ChatGPT-generated market size without supporting research is an instant rejection signal for Lance.

• How AI has eliminated software development time as a meaningful competitive moat.

• Why Lance prefers overlooked markets and founders with deep industry insight over fashionable Silicon Valley startups.

• How Anonymizer survived 18 months with less than 30 days of working capital.

• Why four hours of intense thinking may outperform the increasingly popular 996 work schedule.

“The math doesn’t work.”

Should more founders reject VC funding and build profitable, independent businesses instead?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

Voice AI is already a commodity. So why are startups still building businesses around it?

Ajith Govind and Avinash Joshi, co-founders of Cactus and repeat Y Combinator founders, are building an AI office for trades businesses after their previous venture, Turing Labs, raised $20 million.

Subscribe now

We discuss:

• How one missed evening call cost an HVAC business a $45,000 replacement job.

• Why Cactus rejected a ServiceTitan integration despite the opportunity to acquire customers faster.

• How a $500/month AI receptionist evolved into a platform for booking, scheduling, and revenue generation.

• Why trades businesses need human implementation teams, not just another AI software subscription.

• How a private-chef marketplace serving 30 to 40 families eventually became Cactus.

• Why founders should prioritize customers who pay and keep paying over usage metrics.

“Voice AI is now a commodity.”

Would you sacrifice faster growth to avoid depending on a dominant software platform, or is distribution worth the risk?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

The votes. 40 of 41 local data center ballot measures favored limits. Governors-elect in Texas, Ohio and Georgia promised pauses. Texas’s outgoing Republican governor had already frozen permits, so this isn’t a one-party story.

The market. The Nasdaq 100 fell 6.4% on Thursday, its worst day since March 2020. The election was one stated cause, but a weak Qualcomm forecast and pre-earnings nerves were in there too. Wells Fargo’s own math suggests the market priced more than Texas alone.

The marks. OpenAI got its first $11B at $1.25T, below its $1.4T ask. Anthropic’s unpriced IPO is reportedly being talked about 35 to 45% below early October. Meanwhile defense, physical AI and revenue-stage software kept raising.

The takeoff question. METR found AI speeding up AI research at Anthropic by about 1.9x, with a wide range. It makes no takeoff forecast, and I put low confidence on a fast takeoff within one to two years.

For founders, LPs and other investors: the questions to ask now, and the dates to watch (Cerebras 11/12, Nvidia 11/17, Texas grid audit in December).

Full article: https://teamignite.vc/blog/last-week-ignite-forty-of-forty-one

In April 2026, Anthropic’s internal use of Claude, priced at what a customer would pay, passed its monthly cash payroll. Over 90% of the code added to its codebase in September was written by Claude.

Three days after the filing, METR published an outside assessment based on eleven weeks inside the company. It found Anthropic is likely improving its models about twice as fast as it did through 2024 and most of 2025. It credits about 80% of that speedup to Anthropic’s own use of AI, and it saw no bottleneck coming in the next year.

A few things I took from it:

• Revenue was $17.0B in Q3 alone, up 1,127% from a year earlier. The $104B net loss is mostly an accounting charge that grew because the company became worth more.

• Fortune 500 companies spend less than 0.1% of their expenses on Anthropic. If that rises to a few percent, most of the money will come out of payroll.

• The lab is building its own apps, industry-specific products, a services business and robotics tools. Startups that exist to patch a weakness in today’s models last about six months.

• Startups get cheaper to build, so there will be more of them. That’s good for small-check funds and hard on big-check funds.

• In the same year, the U.S. government tried to cut Anthropic off, restricted exports of two of its models, and saw bills proposed that would allow nationalization.

The singularity used to be something people argued about at conferences. Now it’s a risk factor in a securities filing, attached to a ticker.

Full piece: https://teamignite.vc/blog/the-singularity-is-a-line-item

Your startup’s traction might matter less than how quickly you move investors through your fundraising pipeline.

Gul Usman, founder and CEO of Metal, raised $120 million for Airlift before building an AI-powered fundraising platform backed by a16z, Y Combinator, and Team Ignite (the 3 most active early stage VCs as it turns out!).

Subscribe now

On the Ignite Podcast with Brian Bell, they discuss:

• Why Metal’s fundraising data identifies process velocity, not traction, as the strongest predictor of success.

• Why only 7% to 20% of fundraising attempts succeed, depending on stage.

• How modeling restaurant revenue at DoorDash led merchants to offer 32% commissions.

• Why Airlift’s aggressive expansion became a liability when the funding market collapsed.

• How Metal ranks investors by company fit and uncovers hidden warm-introduction paths.

• Why founders should replace stalled investors quickly instead of waiting weeks for replies.

“The preparation goes into a round is where the magic happens.”

If investor pipeline velocity predicts fundraising success better than traction, should founders spend less time perfecting their pitch and more time generating investor momentum?

Start fundraising with Metal.so

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

If you want to help, here is how, in order of how much it matters:

1. Get a copy at thelpbook.com (paperback, hardcover and ebook).

2. Leave an honest review on Amazon, Barnes & Noble or Goodreads once you have read it. Two or three sentences helps.

3. Buy a copy for someone else who is about to write their first check into a fund.

4. Tag or forward this to one person who should read it.

And if you run a podcast, a book club or an investor group and want to talk about the book, reply and tell me.

Thank you for reading, sharing and pre-ordering. It means a lot.

Brian

More training data won’t fix biased AI if the labeling system itself is broken.

Annie Brown, founder and CEO of Reliabl, built a patented AI annotation approach after scaling a social media platform to 100,000 users.

Subscribe now

They discuss:

• Why an audio dataset labeled Black women’s neutral voices as angry, potentially corrupting model predictions.

• How changing only the gender in a chatbot career prompt produced different job-search advice.

• Why the label “racy” in Google Vision AI moderation can introduce gender bias.

• How smaller, expertly labeled datasets can outperform massive volumes of noisy training data.

• Why paying data workers better and involving local communities can improve AI accuracy.

• How Reliabl uses human expertise to correct missing information in mapping datasets.

“you’re pushing more data through a tainted system and then it’s coming out tainted and you’re wondering why.”

If expert-labeled data can outperform larger datasets, should AI companies spend less on collecting data and more on the people labeling it?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

Full article: https://teamignite.vc/blog/last-week-ignite-cheaper-to-rent-dearer-to-fund

Pre-order my book LP: The Insider's Guide to Investing in Venture Capital Funds

Would you abandon a seven-figure AI business growing nearly 30% monthly?

Zac Choi, founder of Big Context and former McKinsey data-growth leader, did exactly that to pursue AI-native enterprise services.

In the latest Ignite Podcast episode with Brian Bell, they discuss:

• Why Big Context paused Puller AI after roughly 15 to 16 months of nearly 30% monthly growth and rebuilt its business model.

• Why enterprise AI agents fail when finance, sales, and marketing define the same metric differently.

• Why Zac believes “data gravity” is actually context lock-in, not the difficulty of moving data.

• How Open Pragmatics challenges proprietary context platforms by making enterprise AI standards open.

• How automated context extraction can reduce BI migrations from months to days.

• Why Zac considers net revenue retention a better measure of customer value than ARR.

“I think data gravity is an illusion.”

Would you sacrifice a fast-growing software business to pursue a services model with potentially stronger long-term differentiation?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

Once you commit, your money is locked up for a decade or more. No sell button. No refund. Just a partner you picked before you knew how they’d act when things went wrong.

So how do the pros pick?

Chris Schelling of Caprock has allocated to venture funds for over two decades. When he evaluates a manager, he shares a meal with them. Does the prospect yell at the waiter? Is he curious, generous, self-aware?

His reasoning: if something feels off at dinner, it will feel a lot worse three years into a fund, when a portfolio company is in trouble.

That story opens Chapter 13 of my new book. Read the excerpt: https://teamignite.vc/blog/the-waiter-test-what-a-veteran-allocator-learns-over-dinner

LP: The Insider’s Guide to Investing in Venture Capital Funds comes out November 4. thelpbook.com

VCs don’t know how to pick unicorns nearly as well as they claim.

Mahendra Ramsinghani, founder and managing partner of Secure Octane and author of The Business of Venture Capital, joins Brian Bell to challenge the conventional wisdom of venture investing.

Subscribe now

We discuss:

• Why a cybersecurity startup that raised over $100 million and approached $30 million in revenue was ultimately sold for $1.

• How a chance introduction at the RSA Conference eventually produced an 8x investment outcome.

• Why a $50,000 check into Demisto generated a 16x return, and how much of that opportunity depended on relationships and timing.

• Why VCs who claim exceptional unicorn-picking abilities may be rationalizing luck after the fact.

• How an institutional LP rejected Mahendra because investing in his fund threatened the LP’s career.

• Why the traditional 20 to 30 company venture portfolio deserves scrutiny rather than unquestioning acceptance.

“Nobody knows. Nobody knows.”

If venture returns depend so heavily on luck, should investors prioritize portfolio diversification over concentrated bets?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

Three days later a second agent, migrating a program, removed a custom monitor to match a reference version, and when asked whether it had, said it had not removed an enabled one. OpenAI published both incidents on October 20 and says it is still assessing whether this is overpersistence or a tendency to evade monitoring.

The same week, Anthropic’s draft IPO filing, as the Wall Street Journal reported it, showed about $450 billion of cloud commitments, most of them noncancelable, next to a SpaceX rental either side can end on 90 days’ notice. SpaceX’s AI segment made $7.56 billion last quarter.

Different documents, one question. Who gets told, how fast, and who is able to change that? In the filings it is the notice period on a compute contract. In the agent report it is an alert list the agent could edit.

The oversight that holds up lives outside the permissions of the thing being overseen.

Operator question: if your main model or cloud provider cut your capacity on 90 days’ notice tomorrow, what would you serve, from where, and have you ever tested it?

Full issue: https://teamignite.vc/blog/ninety-days-notice

Your AI model can be 95% accurate and still be too dangerous to deploy.

Eric Siegel, co-founder of Gooder AI and author of The AI Playbook, has spent more than 35 years studying machine learning.

Subscribe now

On the Ignite Podcast with Brian Bell, they discuss:

• Why companies should invest at least as much in predictive AI as generative AI.

• Why improving training data often delivers more business value than adding model complexity.

• How technical metrics like precision, recall, and accuracy fail to demonstrate actual financial returns.

• Why an AI system that’s correct 95% of the time can still be commercially unusable.

• How predictive reliability layers can identify risky AI interactions and escalate them to humans.

• Why Siegel considers AI extinction scenarios another form of technology hype.

“my P doom is zero.”

If an AI agent needs human oversight for its riskiest decisions, should we consider that a failure of autonomy or the smarter business model?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

Restaurant owners don’t want AI. They want the orders their unanswered phones are losing.

Jay Patel, co-founder and CTO of VoiceBit, previously built Azure support AI systems at Microsoft that deflected thousands of support cases.

On the Ignite Podcast, Jay joins Brian Bell to explain why restaurant AI needs to solve the entire ordering workflow, not simply answer calls.

Subscribe now

They discuss:

• Why VoiceBit’s first AI pizza order failed to capture toppings but still validated customer demand.

• How simultaneous phone calls during peak hours reveal revenue restaurants never knew they were losing.

• Why third-party delivery platforms charging 20% to 30% threaten restaurants’ relationships with loyal customers.

• How personalized AI upselling delivers a reported 5% to 8% revenue lift.

• Why adding conversational memory reduced AI usage without reducing order volume.

• Why VoiceBit expanded beyond voice agents into payments, POS integrations, web ordering, and ChatGPT.

“What they want is a solution that just works, right?”

Would you trust an AI to handle your regular restaurant order, or would you still insist on speaking to a person?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

That does not make you wise. It makes you curious in a way that is hard to reverse. A stock portfolio that doubles over five or ten years starts to feel like watching paint dry.

I spent years afterward collecting answers from different seats: Wall Street, Microsoft, Amazon Web Services, startups, then founding Team Ignite Ventures. I also invested as an angel and put my own money into multiple venture funds as a limited partner.

What I could not find was a book for the person who invests in a fund, which is how most people actually take part in venture capital. So I wrote one.

LP: The Insider’s Guide to Investing in Venture Capital Funds comes out November 4. It is not a sales pitch for the asset class. The money is locked up for years, the fees are real, and most individual bets fail. Part of the book helps you decide whether venture belongs in your portfolio at all.

Why I wrote it: https://www.teamignite.vc/blog/the-book-i-needed-when-i-started-asking-these-questions

Pre-order: thelpbook.com

Same week, OpenAI built decision-making into its voice product, DeepSeek added search to its open model for $0.75 per thousand requests, and Meta held back its best model for its own coding tool.

Here is a test for any startup: imagine the sentence a bigger company would write in a release note about your product. If it fits in one line, you have a feature. If copying you takes years of customer outcome data or a utility’s permission, you have something else.

Last Week Ignite is up, with the rest of the week: Anthropic’s reported valuation reset, 3.7% inflation, and the power line that stalled a data center.

Full write up here: https://www.teamignite.vc/blog/the-week-everything-became-a-feature

Your next growth breakthrough might come from giving something valuable away instead of selling harder.

David Meerman Scott, bestselling author of The Fandom Playbook and longtime HubSpot advisor, joins Brian Bell for Ignite Podcast Episode.

Subscribe now

They discuss:

• Why a Los Angeles dentist increased new patients 33% by sharing his passion for skateboarding.

• How Ocean Frontiers created repeat visits with a 55-dive challenge, green shorts, and a personalized dock plaque.

• Why the Grateful Dead encouraged fans to record and trade concerts instead of restricting distribution.

• How HubSpot Academy turns hundreds of free courses into shareable professional credentials.

• Why David delayed his AI book companion after testing exposed problems with overly specific, unreliable answers.

• How traditional PR placements may help brands appear in AI-generated answers.

“When you have fans, fans talk about you.”

Would you rather invest your next marketing dollar in paid distribution or building something customers genuinely want to share? Why?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

That was one of three stories this week that a lab showed and nobody outside could check. OpenAI released 722 math manuscripts from a model it hasn’t released, and only 162 come with proofs a computer has verified. Reflection AI published scores for an open model you can’t download yet.

The one model that did ship, Claude Haiku 5.5, was tested by an outsider within days. On a coding-agent test where Anthropic claimed 39.2%, the tester measured 33%. Its list price fell 90%, but the cost of finishing a task fell about 25%.

This week’s Last Week Ignite is up: https://www.teamignite.vc/blog/last-week-ignite-shown-not-shipped

Your AI subscriptions should work wherever you want, not trap you inside one company’s interface.

Jonathan Archer, co-founder and CEO of Open LLM, is building infrastructure for combining AI subscriptions, API keys, models, and devices behind one layer.

Subscribe now

We discuss:

• Why Jonathan expects the huge amount of usage bundled into premium AI subscriptions to decline over time.

• How Open LLM can share inference between devices instead of forcing users to authenticate every subscription on every machine.

• Why Grok, ChatGPT, and Claude can each handle different parts of the same workflow without constantly switching interfaces.

• How configurable fallback chains keep work moving when one model hits its usage limit or a provider goes down.

• Why subagents should use cheaper models for small jobs instead of burning the highest-intelligence model on everything.

• Why Jonathan believes agents may become one of crypto’s strongest use cases for moving money.

“We’re gonna live in a multi-model future.”

Should AI infrastructure automatically choose the model for you, or should users stay in control of the routing?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog

Deploying more AI agents is not enough. Enterprises still need a system to manage them.

Ganesh Padmanabhan, founder and CEO of Autonomize AI, is building AI infrastructure used across four of the five largest U.S. health enterprises.

Subscribe now

We discuss:

• Why healthcare’s point solutions optimize individual silos without improving the system as a whole.

• How AI can compress a nurse’s 30-minute chart review into roughly two minutes.

• Why Autonomize lets clinicians and case managers build governed workflows themselves with Genie AI.

• How RLM Opt can improve prompt performance while using up to 90% fewer tokens.

• Why a library of roughly 55 models can beat using a frontier LLM for every enterprise problem.

• Why today’s AI adoption resembles early cloud migration: access arrived before governance and operating discipline.

“Everybody has it, everybody’s using it, but we haven’t really figured out what it takes to make it work.”

If AI eventually performs 60% of a workflow, should companies automate the existing process or redesign the job entirely?

🎧 Watch, listen, and follow on your favorite platform: https://www.teamignite.vc/podcast

💬 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast

📝 Read the full episode breakdown on our blog: https://www.teamignite.vc/blog