Ваня Замесин

Автор методологий Next Move Theory и Advanced JTBD Author | 14k студентов

[Ваня Замесин @ivanzamesin]

@aleks_shamis could you share the prompt and general approach please?

[Ваня Замесин @ivanzamesin]

@mikeboysen 100% it had to be done

[Ваня Замесин @ivanzamesin]

JTBD is widely misunderstood because it was never connected to value creation, unit economics, assumption testing and product strategy. Let's fix that. https://t.co/ajpIou7KtQ

[Aleks Shamis @aleks_shamis]

About six months ago we set a really cool, ambitious goal for ourselves. Within a year, the only thing we'll do is talk. Talk to founders, talk among ourselves, talk to experts, partners, and sometimes talk to AI. It's a big task to hand AI so many operations. We know that from everything we do: take notes, organize notes, write memos, build reporting tables, answer emails, make introductions, write newsletters, organize events, write LP reports, go through audits, etc. The only time we deliver strategic value that only we can deliver is when we talk to people. The rest is necessary, not always pleasant, and demands excellent execution, but it all stems from the quality of the conversations we have.

Man, am I glad we built https://t.co/ZVERTWpOoV. The boring operational part of my job turned into a fun game of destroying manual routines, one after another, with a tool that can do virtually anything, seamlessly, collaboratively, with a great UX. Even small things like writing and posting on social media have become a joy.

[Ваня Замесин @ivanzamesin]

Google isn’t SOTA by choice — Flash is tuned for speed and cost because AI has to stay profitable across Search, Gmail, YouTube, and Android massive scale

[Ваня Замесин @ivanzamesin]

@nicklaunches Skills that do product strategy and general product work better than 95% of PMs

https://t.co/mKoAzL5EX8

[Ваня Замесин @ivanzamesin]

@Joi2James Skills that do product strategy and general product work better than 95% of PMs

https://t.co/mKoAzL5EX8

[Ваня Замесин @ivanzamesin]

@volosatovde Skills that to product work better than 95% of PMs

https://t.co/mKoAzL5EX8

[Ваня Замесин @ivanzamesin]

@billgnofficial Skills that to product work better than 95% of PMs

https://t.co/mKoAzL5EX8

[Ваня Замесин @ivanzamesin]

@anupamrjp Skills that to product work better than 95% of PMs

https://t.co/mKoAzL5EX8

[Ваня Замесин @ivanzamesin]

@lightsilver323 Skills that do market research, write PRDs and act like an experienced PM

https://t.co/mKoAzL5EX8

[Ваня Замесин @ivanzamesin]

@lightsilver323 Skills that do product strategy and daily product work better than 90% of PMs

https://t.co/mKoAzL5EX8

[Ваня Замесин @ivanzamesin]

I strongly believe ultrafast is the next step change for LLMs, because you’ll be able to work at the speed of thought without switching context while waiting for the agent. We’ll stay in flow for hours and create at incredible speed.

[Ваня Замесин @ivanzamesin]

“In the beginning was the Word.” — John 1:1

So God used an LLM. Spoke the universe into existence and rested on day seven because He got rate limited.

[Ваня Замесин @ivanzamesin]

Selling tokens is a bad business model long-term, because clients don’t care about tokens. They care about getting the job done cheaply, quickly, and reliably. This is why Opus 4.5 was so good

[Ваня Замесин @ivanzamesin]

Dario after the Fable release drama, looking at Kimi 3 and Qwen 3.8 benchmarks

[Paul Mit @pmitu]

Most startup advice is garbage.

🍊 YCombinator's isn't.

23 YC lessons worth stealing:

• Launch ASAP. Ideally today.
• Build what users want, not what your imagination wants.
• Do things that don't scale before PMF.
• Write code. Talk to users. Repeat.
• Find 10 people who absolutely love your product.
• Then find 50 more.
• Don't scale team, marketing, or product before real PMF.
• Until PMF: stay small, fast, and flexible.
• Valuation ≠ success.
• Focus. Startups can only solve one big problem at a time.
• Founder relationships matter more than most people realize.
• Don't freak out. Every startup hits rough patches.
• Organic growth is what happens when a great product meets the right market.
• Avoid long enterprise sales cycles if you can.
• Avoid becoming a custom dev shop for corporations.
• Avoid conferences unless they reliably bring customers.
• Fire customers who can sink the ship.
• Watch competitors. Don't blindly copy them.
• Most startups don't die because they run out of money 🤷‍♂️
• Don't be assholes.
• Sleep. Exercise. Founder energy compounds.
• Use 90/10 thinking: get 90% of the outcome with 10% of the effort.
• VC money isn't your money. Don't be afraid to spend it—but maybe take it easy on the escorts and sports cars.

YC is full of smart people after all.

Worth paying attention.

[Ваня Замесин @ivanzamesin]

A founder has to live torn between two poles: infinite belief in their product (without it, you can't raise money or attract talent) and the paranoia that nobody needs it (because there's 10 hidden risks that will kill it).

The real work happens between these two poles.

[Ваня Замесин @ivanzamesin]

Founders don't skip the fundamentals because they don't know them.

I burned about $300K of my own money learning this.

Contact with reality hurts. That's the entire reason people postpone it. That's the reason I postponed it.

It's also the entire game. The founders winning right now vibe-code something ugly in a week and go try to sell ASAP. They're buying the exact location of their own error, while it's still cheap to fix.

Revenue is a side effect of the search.

Writing a great story about a huge market and remarkable technology is easy. A smart founder does it in an afternoon and it sounds plausible.

But seasoned investors will drill you about fundamentals.

Getting to sales that repeat and scale is brutally hard. It takes longer than you planned. It costs more than you modelled. And it will tell you things about your idea you did not want to hear.

[Ваня Замесин @ivanzamesin]

Just a week ago it was all drama: "OMG, Fable 5 is so dangerous, regulate it NOW! Oh, they're regulating it! Wow!" Now Sol and Kimi K3 are out, matching or beating Fable 5.

I'm wondering what PR trick Anthropic cooks up for the next models. What tops Fable 5 drama?