Founders · Wednesday, 26 August 2026 · Issue 85 · 10 stories

Founders are running from their AI agents

The bots were supposed to free you up. Instead, some founders now log 90-hour weeks keeping pace with their own fleets, and that's just one story in a day when the real signal is about who's actually funding what, and why.

Funding & VC

Ventures Platform's $84M fund pushes past Nigeria into four African markets

Ventures Platform closed an oversubscribed $84 million Fund II, nearly doubling its first fund and leaving Nigeria behind: it has already cut checks to five startups in Kenya, South Africa, and Egypt. Checks now run up to $3 million, with an AI thesis focused on rewriting cost structures for African markets. The firm points to a harsher LP environment that rewards capital efficiency and realized returns, not just successive raises.

techcrunch.com ↗4 min read at source

Robot-brain startup Generalist hits $3B in a frantic bet on a "ChatGPT moment"

Generalist, the quiet two-year-old building a foundation model for all robots, is now valued at $3 billion after a ~$200M extension led by 8VC, growing its Series B from $400M at June's $2B mark to $600M total. Its Gen 1.5 model already teaches robots new tasks from 3-to-12-second video demos. Rivals Physical Intelligence ($11B) and Skild ($14B) show investors betting a general robot brain is near, though some VCs warn it's years off.

techcrunch.com ↗2 min read at source

Stability AI locks in $76M from UMG, Sony, Warner and EA

Stability AI banked $76M in new Series B funding, lifting its total haul to $232M. The unusual investor roster, Universal Music, Sony Music, Warner Music and EA, alongside AMD Ventures, reflects how the round is really about co-development deals, not just capital. Music and gaming giants are effectively betting on building Stability's generative tools into their own creative workflows.

techcrunch.com ↗1 min read at source

Seed rounds rarely die from bad ideas. They die from these 8 process mistakes.

Most seed raises fail before the term sheet, not because the idea is weak but because the process is broken. The biggest self-inflicted wound: an arbitrary ask size that signals a founder has no real plan. Fix it with a bottoms-up budget tied to 18-24 months of runway and what gets you to Series A. Investors are underwriting execution, and a sloppy fundraise is your first test.

startupfundraising.com ↗12 min read at source

Runable raises $21M to chase the harder half of AI: winning customers, not just building apps

Bengaluru's Runable started as an AI scraping play, pivoted into a general-purpose agent, then realized the real pain point was downstream. Users could build sites and slide decks in minutes, but getting customers still meant juggling ad accounts, SEO, and chatbots. Its $21M Series A funds that "grow" agent, targeting small businesses instead of fighting OpenAI and Cursor on the building turf. The bet: hit $2M ARR three weeks after launch suggests paying for outcomes, not tools.

techcrunch.com ↗5 min read at source

VC's software-era metrics break down on capital-heavy energy startups

Dominant VC math is tuned for software: near-zero marginal cost scaling, recurring revenue, five-to-tenfold annual growth. Energy and materials startups live under different physics, where capital arrives early, stays long, and doubling revenue is often the ceiling. Their real traction signals are reaction yield, product stability, and customer qualification, not user curves. Investors who score them on ARR routinely demand impossible speed.

cen.acs.org ↗4 min read at source

Growth & GTM

Solo dev let Reddit design his car games; app hit #6 in paid Travel

One engineer, a solo dad of five, built Pit Stop: a $1.99 app that maps every business at all 19,675 US interstate exits, fully offline. He asked two subreddits what games families play in the car; by midnight both were live, including Cow Collector, which only unlocks when GPS passes a real cemetery (4,644 pulled from OpenStreetMap). Next day the app hit #6 in paid Travel, and shipping became his flywheel: ask Reddit, ship in 24 hours, turn the data into content and press pitches.

pit-stop-app.com ↗4 min read at source

Leadership & Lessons

AI Agents Were Supposed to Ease Founders' Load. Now They're Pulling All-Nighters.

The AI agents sold to founders as a cure for burnout are now their most demanding bosses. Keel co-founder Aditya Sharma went to bed at 6 a.m. after an all-nighter spent supervising bots that build features, run research, and watch marketing. The productivity FOMO cuts hard against the industry claim that automation eases the load. Some founders now log 90-hour weeks just keeping pace with their own fleets.

wsj.com ↗3 min read at source

Market & Trends

Second place decayed: BigCommerce's 19x revenue gap to Shopify is now 44x

Six years after turning down a $1.5B Intuit bid to IPO for a $4.8B market cap, BigCommerce (now Commerce.com) trades under 1x ARR and is fielding a hostile bid. The revenue gap to Shopify swelled from 19x to 44x, not on collapse but on pure deceleration: 44% growth to negative, while a company 20x its size held steady at roughly 26-30%. The B2B wedge failed too, as Shopify's B2B volume grew 76% to BigCommerce's 17%. A sharp, data-dense case study on why second place doesn't compound.

saastr.com ↗8 min read at source

Per-seat pricing is dying. Here are the 3 models fighting to replace it

ZoomInfo CEO Henry Schuck says the seat model is running out of road, and the evidence is brutal: four straight years of 12%+ SaaS inflation have emptied buyers' patience, and 45% of CIOs now fund AI tokens straight out of existing software budgets. Three new pricing models are vying to replace seats, but even the top consultants can't agree on the winner. The deciding test: can you name a single countable thing your software does a customer would pay for on its own?

saastr.com ↗16 min read at source

You’re caught up.

That was all 10 stories in the 26 August issue.