Y CombinatorLightcone: Consumer is back, What’s getting funded now, The vibes immaculate
CHAPTERS
- 0:00 – 1:51
Why this YC batch feels different: energy, growth, and a platform reset
The partners open by noting unusually high momentum in the Winter 2024 batch and tease the core explanation: a major platform shift. They set up the episode’s goal—using YC’s internal batch stats and qualitative observations to explain what’s changing and why it matters for founders and investors.
- •Winter 2024 batch has notably strong “vibes” and execution pace
- •AI-driven platform shift makes old assumptions about software markets feel reset
- •Batch-wide growth highlighted as a sign of real momentum
- •Episode will compare W24 data to earlier batches (e.g., W20)
- 1:51 – 2:40
AI dominates the batch: from 8% in W20 to ~70% now
They quantify how central AI has become: roughly 70% of the batch is building with AI, a dramatic jump from a few years earlier. They contrast today’s AI wave with earlier “machine learning” eras and discuss how being early (e.g., Replicate) paid off over time.
- •~70% of W24 companies are AI (~170 companies)
- •W20 had ~8% AI; shift is rapid and massive
- •Earlier AI builders faced slow adoption due to small market
- •Terminology and market maturity changed from ‘ML’ to mainstream ‘AI’
- 2:40 – 5:21
Consumer is back: pivots toward consumer and debate over ‘tarpit’ ideas
Harj observes a resurgence of consumer startups and founders pivoting into consumer during the batch. The group debates whether this is a positive renaissance or a risky return to easy-but-hard-to-win consumer “tarpits,” with examples of classic traps like travel planning.
- •More founders pivoting into consumer than in recent years
- •Risk: consumer tarpit ideas feel easy but are difficult to scale
- •Counterpoint: consumer products are fun and historically drove YC’s early era
- •Shift away from ‘only B2B SaaS is viable’ mindset
- 5:21 – 6:45
Fear of incumbents (Facebook then, OpenAI now) and why there’s still white space
They revisit why consumer slowed down—platform incumbents like Facebook ‘sucking oxygen’ and cloning features. They map that concern onto today’s AI landscape (OpenAI as the new gravitational center) while arguing expanding capabilities create too many niches for one player to own.
- •Past consumer pessimism driven by fear of Facebook copying/crushing startups
- •Modern analog: foundational model providers could become the new gatekeepers
- •Argument for optimism: capability expansion creates too many opportunities to cover
- •Timing matters—incumbent ‘crushing’ hasn’t fully played out yet
- 6:45 – 10:03
Developer tools surge: infrastructure first, apps later
Diana explains why dev tools are up sharply: AI is creating new engineering patterns and a lot of repeated ‘plumbing’ work. They frame this as a two-phase technology cycle—first build infrastructure (rails), then see mass application proliferation.
- •W24 funded ~30% more dev tools than four years ago
- •AI app building requires heavy plumbing (RAG, indexing, evals, fine-tuning)
- •Tooling phase precedes a broader wave of applications
- •Potential long arc from software tooling to even custom silicon
- 10:03 – 11:23
Open source dev tools and how they’re judged: mindshare, adoption, and Hacker News
They discuss open source as a powerful go-to-market and adoption strategy for dev tools, where early success looks more like consumer growth than revenue. Signals include GitHub stars, adoption by tastemakers, and production usage—often catalyzed by a strong Hacker News launch.
- •Early-stage OSS often prioritizes adoption over immediate monetization
- •Key signals: GitHub star growth, tastemaker endorsement, early production usage
- •Hacker News functions like a distribution channel for dev tools
- •Dev tools resemble ‘consumer-style’ marketing to a ~20M developer audience
- 11:23 – 13:37
Supabase case study: OSS challenger playbook and internal batch adoption
Using Supabase as an example, they show how an open source competitor can attack a closed-source or incumbent platform and win mindshare. They highlight Supabase’s breakout via Hacker News and note its remarkable adoption inside the current YC batch.
- •Supabase launched as OSS alternative to Firebase and grew via Hacker News
- •Reading HN comments can reveal product-market pull from elite developers
- •~73 companies (about a third of the batch) use Supabase infrastructure
- •Broader pattern: open source challengers vs incumbents (GitLab, Mattermost, Posthog)
- 13:37 – 14:04
The most technical batch ever: why technical founders and younger founders increased
They share a striking stat: nearly every company has a technical founder, and the median founder age is lower than a few years ago. The group attributes this to AI being cutting-edge and favoring builders without legacy ‘baggage,’ plus the renewed belief that this is a rare moment to build.
- •~99% of companies have a technical founder (vs ~88% during pandemic)
- •Median founder age shifted younger (about 30 → 26)
- •AI raises the bar: technical capability is table-stakes
- •More students consider dropping out due to perceived once-in-a-lifetime window
- 14:04 – 20:31
From ‘software eats the world’ to ZIRP-era tech-enabled bets—and back to pure software
They unpack how venture dollars drifted toward ‘tech-enabled’ businesses when core software opportunities felt saturated, especially during the ZIRP era. Now, AI creates a new platform shift that pulls capital and talent back toward high-leverage software opportunities.
- •Tech-enabled wave grew as VCs sought places to deploy capital without new platforms
- •Examples span Flexport (works) to WeWork (claimed tech-enabled, failed)
- •AI represents a new platform shift with higher ROI than tech-enabled real estate/ops plays
- •Key nuance: gross margin matters more than labels for business quality
- 20:31 – 24:11
What YC funded less of: local-market clones, international copies, and marketplaces
They explain the decline in startups copying US models into other countries (and local-market plays more broadly). As these opportunities mature, the batch becomes more US-centric and less marketplace-heavy, reflecting that the best new opportunities now feel global by default (especially AI/dev tools).
- •Less ‘DoorDash for X’ / ‘Robinhood for Y’ style company formation
- •International share drops (about 45% in W20 → ~25% in W24)
- •Marketplace ideas down sharply (noted as ~4x less than 2020)
- •AI/dev tools tend to be global products, not local-market businesses
- 24:11 – 31:51
Crypto fades despite a bull run: mindshare shift, trauma, and regulatory chill
They note the surprising lack of crypto companies in W24 even as Bitcoin rises again. The partners attribute it to AI absorbing builder attention, the hangover from failed crypto startups, and the fear created by US regulation-by-enforcement.
- •Crypto applications did not rebound with the latest Bitcoin price surge
- •AI now dominates engineer mindshare that crypto once captured
- •Post-crash ‘startup stigma’ discouraged some former crypto founders
- •US regulatory enforcement creates fear of severe consequences for success
- 31:51 – 37:09
Products feel real again: demo culture, Product Day, and returning to YC’s builder roots
They describe internal initiatives that emphasize product craftsmanship: on-stage Product Day demos and live Bookface launch events. The partners compare the feeling to earlier YC eras when teams pushed technical boundaries weekly, and highlight how AI demos can feel like ‘homebrew computer club’ moments.
- •Product Day showcased tangible, working demos; lots of ‘wow’ moments
- •Examples: AI software engineer implementing dark mode; voice agents nearing Turing-level realism
- •Bookface live demos rewarded strong internal launches and encouraged deep technical Q&A
- •Cultural shift back from ‘growth and sales’ toward invention and product focus
- 37:09 – 38:50
AI companies scaling revenue fast: batch ARR triples from January to Demo Day
They argue this wave is not just hype: many companies quickly reach real recurring revenue by selling to businesses and replacing labor costs, not merely reallocating software budgets. They cite a striking aggregate metric—batch ARR tripling over the program.
- •~80% of companies entered with no revenue; many hadn’t fully launched
- •Total batch ARR grew from ~$6M (Jan) to ~$20M (Demo Day)
- •AI enables monetization via labor budget replacement, not only SaaS spend
- •Growth suggests non-zero-sum value creation and compounding impact
- 38:50 – 40:33
Investor optimism and ‘everything is up for grabs’: the SaaS reset thesis
They describe the in-person investor reception and how it signaled a broad reset in what investors consider fundable. With AI, previously ‘unfundable’ categories—like direct Salesforce competitors—now seem plausible because the platform itself is changing.
- •In-person investor reception created concentrated momentum and deal flow
- •Investors revisiting old taboos (e.g., funding Salesforce competitors)
- •Example: ‘AI Salesforce’ rebuild attracting strong interest
- •Core claim: platform shift puts every SaaS dollar back in contention
- 40:33 – 42:21
Looking ahead: more pivots, longer cycles than people expect, and why it’s early
They close by arguing the opportunity is far from exhausted: the batch had unusually high pivot rates and the platform shift is still in early innings. They compare today to earlier tech waves where the biggest outcomes arrived years later, and encourage founders to apply for the next batch.
- •High pivot rate: ~30% in W24 vs ~10% four years ago
- •AI era likened to early social/web moments—big winners may be years away
- •Trends play out longer than expected (Airbnb/DoorDash/Coinbase came later)
- •Call to action: YC applications open; now is a strong moment to start
