Y Combinator
About this creator
Y Combinator publishes practical startup guidance on product building, fundraising, founder decision-making, and scaling high-growth technology companies.
Why they're in the library
Included for practical education on practical startup guidance on product building, fundraising, founder decision-making, and scaling high-growth technology companies.
Showing 320 digests for Y Combinator.
AI strategy & adoption
YouTube
How fluid intelligence and program search point toward AGI
Y Combinator
July 3, 2025
AI strategy & adoption
9
MVP & prototyping
5
50,000x more compute barely moved ARC scores — scaling alone fails.
Test-time adaptation unlocked fluid intelligence; it is still not enough.
AGI needs deep learning intuition fused with discrete program search.
AI strategy & adoption
YouTube
Fei-Fei Li on spatial intelligence as the next frontier in AI
Y Combinator
July 1, 2025
AI strategy & adoption
9
Founder interviews
7
AGI cannot be complete without solving 3D spatial intelligence.
Vision took 540M years to evolve — harder than language by design.
Intellectual fearlessness is the single trait that predicts success.
AI strategy & adoption
YouTube
How consumer AI will evolve: product, distribution, and emotional relationships
Y Combinator
June 27, 2025
AI strategy & adoption
9
Product-market fit
7
Growth hacking
6
AI is shifting consumers from attention-based to relationship-based products
Great products beat marketing tricks — there is no distribution shortcut
Health and personal security are the highest-opportunity categories for consumer AI
AI strategy & adoption
YouTube
Satya Nadella on AI strategy, compute, and the future of software
Y Combinator
June 25, 2025
AI strategy & adoption
9
Founder interviews
6
Management
5
Change management, not compute, is the biggest blocker to AI deployment
Software engineering abstracts upward — the job becomes software architect
AI needs social permission: prove real-world surplus or lose the mandate
Founder interviews
YouTube
Sam Altman on OpenAI's origins, AI agents, and building for the future
Y Combinator
June 21, 2025
Founder interviews
10
AI strategy & adoption
8
Niche selection
6
Massive product overhang exists — reasoning models are barely being used yet.
Don't build a ChatGPT clone; unexplored ground is where startups win.
AI for science is the highest-leverage bet for improving human lives.
AI tools & automation
YouTube
Software is changing again: Karpathy on LLMs, agents, and building for the AI era
Y Combinator
June 19, 2025
AI tools & automation
9
AI strategy & adoption
9
MVP & prototyping
6
LLMs are operating systems, not utilities — we're in the 1960s
The generation–verification loop is the bottleneck every product must optimise
Agents can't click: build your docs and APIs for LLMs now
Founder interviews
YouTube
Elon Musk on digital superintelligence, multiplanetary life, and being useful
Y Combinator
June 19, 2025
Founder interviews
9
AI strategy & adoption
7
Digital superintelligence — smarter than any human — arrives within two years.
Truth-seeking is the single most critical property for AI safety.
Ego above ability breaks your feedback loop to reality and kills companies.
Founder interviews
YouTube
Alexandr Wang on building Scale AI, agentic work, and competing with China
Y Combinator
June 18, 2025
Founder interviews
10
AI strategy & adoption
9
Pivoting
8
Every firm's future moat is a fine-tuned model on its own data.
China likely leads on AI data; US grid capacity is a critical weakness.
Human workers will gain the infinite-leverage boost programmers already have.
Founder interviews
YouTube
Cursor CEO Michael Truell on the future of software development
Y Combinator
June 11, 2025
Founder interviews
9
AI strategy & adoption
7
Taste — knowing what to build — is the one skill AI cannot automate.
Cursor's moat mirrors search: scale generates signal that sharpens the model.
Building an editor, not an extension, was the non-obvious bet that paid off.
Prompt engineering
YouTube
State-of-the-art prompting and evals for AI agents
Y Combinator
May 30, 2025
Prompt engineering
10
AI strategy & adoption
7
B2B sales
5
Your evals are the crown jewels — not your prompts.
Meta-prompting lets one prompt generate better versions of itself.
Forward-deployed engineers win enterprise deals by demoing, not selling.
AI strategy & adoption
YouTube
How to design AI-native apps by exposing the system prompt to users
Y Combinator
May 23, 2025
AI strategy & adoption
9
Design thinking
7
Business models
5
Hidden system prompts make AI generic — letting users edit them fixes this.
Tools transform LLMs from chatbots into agents that complete work autonomously.
Every domain will have its Cursor moment when users can teach AI their workflows.
AI strategy & adoption
YouTube
AI unlocks startup ideas that previously failed on economics or timing
Y Combinator
May 16, 2025
AI strategy & adoption
9
Business models
8
Product-market fit
5
Old three-sided marketplaces collapse to two-sided when AI handles evaluation
Full-stack startups failed on gross margins — AI agents finally fix that
Follow the technology first; lean startup advice is now outdated
AI strategy & adoption
YouTube
How AI coding agents will reshape software engineering and knowledge work
Y Combinator
May 9, 2025
AI strategy & adoption
9
Business models
6
A non-engineer shipped 35,000 lines of production code writing none of it
Software engineering jobs as we know them will not exist in a decade
Now is the best moment in history to start a company from scratch
Founder interviews
YouTube
Windsurf CEO: how a GPU startup pivoted twice to build an AI coding tool
Y Combinator
May 2, 2025
Founder interviews
9
Pivoting
8
AI strategy & adoption
6
Scrapped a $2M GPU business in one weekend to chase coding AI
Every competitive insight depreciates — continuous new bets are the only moat
Non-technical users now build production tools without ever opening the editor
Practical vibe coding techniques for better AI-generated code
Y Combinator
April 25, 2025
Processes & SOPs
9
AI tools & automation
8
Git reset beats prompting repeatedly — clean slate fixes accumulate less junk.
Write high-level integration tests before moving to the next feature.
Mature frameworks like Rails produce more reliable LLM output than niche languages.