Parul Singh
@parulia.bsky.social
board partner @ Initialized. cofertility, stellarsleep, medivis, mantle bio. SWE & founder --> investor.
🧵 The stickiest AI products don’t feel magical.
They feel inevitable.
They don’t just show off what the model can do.
They erase something frustrating from your workflow—so smoothly, you can’t imagine going back.👇
They feel inevitable.
They don’t just show off what the model can do.
They erase something frustrating from your workflow—so smoothly, you can’t imagine going back.👇
July 8, 2025 at 2:08 PM
🧵 The stickiest AI products don’t feel magical.
They feel inevitable.
They don’t just show off what the model can do.
They erase something frustrating from your workflow—so smoothly, you can’t imagine going back.👇
They feel inevitable.
They don’t just show off what the model can do.
They erase something frustrating from your workflow—so smoothly, you can’t imagine going back.👇
Reposted by Parul Singh
1/
🔗 Founders are sleeping on one of the most powerful uses of AI right now:
Prompt chains. 🔗
Not for writing emails or fixing grammar.
But for thinking. Strategically.
Here’s how I used one to generate startup ideas, GTM plans, SEO strategy—and more—in under an hour 🧵
🔗 Founders are sleeping on one of the most powerful uses of AI right now:
Prompt chains. 🔗
Not for writing emails or fixing grammar.
But for thinking. Strategically.
Here’s how I used one to generate startup ideas, GTM plans, SEO strategy—and more—in under an hour 🧵
June 23, 2025 at 6:05 PM
1/
🔗 Founders are sleeping on one of the most powerful uses of AI right now:
Prompt chains. 🔗
Not for writing emails or fixing grammar.
But for thinking. Strategically.
Here’s how I used one to generate startup ideas, GTM plans, SEO strategy—and more—in under an hour 🧵
🔗 Founders are sleeping on one of the most powerful uses of AI right now:
Prompt chains. 🔗
Not for writing emails or fixing grammar.
But for thinking. Strategically.
Here’s how I used one to generate startup ideas, GTM plans, SEO strategy—and more—in under an hour 🧵
Name a founder more people should be learning from—someone whose name doesn’t get mentioned enough.
I’ve been lucky to back and learn from some remarkable founders. The ones I keep thinking about aren’t always the loudest or most followed. (1/3)
I’ve been lucky to back and learn from some remarkable founders. The ones I keep thinking about aren’t always the loudest or most followed. (1/3)
June 12, 2025 at 4:40 PM
Name a founder more people should be learning from—someone whose name doesn’t get mentioned enough.
I’ve been lucky to back and learn from some remarkable founders. The ones I keep thinking about aren’t always the loudest or most followed. (1/3)
I’ve been lucky to back and learn from some remarkable founders. The ones I keep thinking about aren’t always the loudest or most followed. (1/3)
If AI could take one thing off your plate today -- what would you hand over without hesitation?
I’ve been asking this in 1:1s with founders. And what’s surprising isn’t that they’re automating more -- it’s what they’re automating first. (1/3)
I’ve been asking this in 1:1s with founders. And what’s surprising isn’t that they’re automating more -- it’s what they’re automating first. (1/3)
June 11, 2025 at 6:20 PM
If AI could take one thing off your plate today -- what would you hand over without hesitation?
I’ve been asking this in 1:1s with founders. And what’s surprising isn’t that they’re automating more -- it’s what they’re automating first. (1/3)
I’ve been asking this in 1:1s with founders. And what’s surprising isn’t that they’re automating more -- it’s what they’re automating first. (1/3)
Everyone’s watching AI demos.
But the real AI startups? They’re compounding quietly.
I went deep on the Lean AI Leaderboard — dozens of startups with breakout traction.
Here are 10 patterns from the teams actually pulling ahead 🧵
But the real AI startups? They’re compounding quietly.
I went deep on the Lean AI Leaderboard — dozens of startups with breakout traction.
Here are 10 patterns from the teams actually pulling ahead 🧵
June 10, 2025 at 2:13 PM
Everyone’s watching AI demos.
But the real AI startups? They’re compounding quietly.
I went deep on the Lean AI Leaderboard — dozens of startups with breakout traction.
Here are 10 patterns from the teams actually pulling ahead 🧵
But the real AI startups? They’re compounding quietly.
I went deep on the Lean AI Leaderboard — dozens of startups with breakout traction.
Here are 10 patterns from the teams actually pulling ahead 🧵
None of these existed in 2020.
Anthropic: $61.5B
Cohere: $5.5B
ElevenLabs: $3.15B
Harvey: $5B
Moonshot AI: $3B
Together AI: $3.3B
Figure AI: $39.5B
Crusoe: $2.8B
Character.AI: $10B
Perplexity AI: $18B
Adept AI: $20B
Mistral AI: $6.2B
Combined valuation: $178 billion. #MondayMotivation
Anthropic: $61.5B
Cohere: $5.5B
ElevenLabs: $3.15B
Harvey: $5B
Moonshot AI: $3B
Together AI: $3.3B
Figure AI: $39.5B
Crusoe: $2.8B
Character.AI: $10B
Perplexity AI: $18B
Adept AI: $20B
Mistral AI: $6.2B
Combined valuation: $178 billion. #MondayMotivation
June 9, 2025 at 3:07 PM
None of these existed in 2020.
Anthropic: $61.5B
Cohere: $5.5B
ElevenLabs: $3.15B
Harvey: $5B
Moonshot AI: $3B
Together AI: $3.3B
Figure AI: $39.5B
Crusoe: $2.8B
Character.AI: $10B
Perplexity AI: $18B
Adept AI: $20B
Mistral AI: $6.2B
Combined valuation: $178 billion. #MondayMotivation
Anthropic: $61.5B
Cohere: $5.5B
ElevenLabs: $3.15B
Harvey: $5B
Moonshot AI: $3B
Together AI: $3.3B
Figure AI: $39.5B
Crusoe: $2.8B
Character.AI: $10B
Perplexity AI: $18B
Adept AI: $20B
Mistral AI: $6.2B
Combined valuation: $178 billion. #MondayMotivation
Everyone jokes about the 10X engineer.
But the best ones?
They absolutely exist—and they’re evolving.
🧵 The New Engineering Stack in the age of AI:
But the best ones?
They absolutely exist—and they’re evolving.
🧵 The New Engineering Stack in the age of AI:
June 3, 2025 at 5:30 PM
Everyone jokes about the 10X engineer.
But the best ones?
They absolutely exist—and they’re evolving.
🧵 The New Engineering Stack in the age of AI:
But the best ones?
They absolutely exist—and they’re evolving.
🧵 The New Engineering Stack in the age of AI:
VC is broken. AI killed the old playbook. Here’s what’s next 🧵
April 15, 2025 at 2:03 PM
VC is broken. AI killed the old playbook. Here’s what’s next 🧵
Maybe we should all post inventions and innovations for a week, “made possible by NIH” or other grants.
Science is historically not good enough at self promotion and this is a storytelling war.
Science is historically not good enough at self promotion and this is a storytelling war.
The tiniest ever, bioresorbable pacemaker, activated wirelessly by light, and its potential as a platform for electrotherapy
nature.com/articles/s41...
from work supported by NIH
nature.com/articles/s41...
from work supported by NIH
April 2, 2025 at 3:57 PM
Maybe we should all post inventions and innovations for a week, “made possible by NIH” or other grants.
Science is historically not good enough at self promotion and this is a storytelling war.
Science is historically not good enough at self promotion and this is a storytelling war.
Reposted by Parul Singh
@parulia.bsky.social distills to the core
In a world where software is a commodity and GPT-4 is table stakes…
What actually gives startups a moat?
Let’s talk about two of the most powerful—but often misunderstood—defenses:
Data moats vs. Brand moats.
🧵
What actually gives startups a moat?
Let’s talk about two of the most powerful—but often misunderstood—defenses:
Data moats vs. Brand moats.
🧵
April 2, 2025 at 11:54 AM
@parulia.bsky.social distills to the core
In a world where software is a commodity and GPT-4 is table stakes…
What actually gives startups a moat?
Let’s talk about two of the most powerful—but often misunderstood—defenses:
Data moats vs. Brand moats.
🧵
What actually gives startups a moat?
Let’s talk about two of the most powerful—but often misunderstood—defenses:
Data moats vs. Brand moats.
🧵
April 2, 2025 at 9:00 AM
In a world where software is a commodity and GPT-4 is table stakes…
What actually gives startups a moat?
Let’s talk about two of the most powerful—but often misunderstood—defenses:
Data moats vs. Brand moats.
🧵
What actually gives startups a moat?
Let’s talk about two of the most powerful—but often misunderstood—defenses:
Data moats vs. Brand moats.
🧵
Embedded AI’s quietly winning—like spellcheck sneaking into Word and ruining standalone’s day. Thought UX didn’t matter? It’s everything now. Standalone apps are out there fighting gravity.
April 1, 2025 at 3:19 PM
Embedded AI’s quietly winning—like spellcheck sneaking into Word and ruining standalone’s day. Thought UX didn’t matter? It’s everything now. Standalone apps are out there fighting gravity.
AI agents are here—CB Insights shares the wild stat that 60% of AI startups in 2025 are building self-running systems. Why now:
✅ Agents handle tasks solo
🤖 Scale without humans
🏎 Outpace LLMs fast
✅ Agents handle tasks solo
🤖 Scale without humans
🏎 Outpace LLMs fast
March 31, 2025 at 3:19 PM
AI agents are here—CB Insights shares the wild stat that 60% of AI startups in 2025 are building self-running systems. Why now:
✅ Agents handle tasks solo
🤖 Scale without humans
🏎 Outpace LLMs fast
✅ Agents handle tasks solo
🤖 Scale without humans
🏎 Outpace LLMs fast
VC myths are killing your raise—here’s the truth:
📈 Raw traction beats polish (Notion’s buzz)
❤️ User love outpitches you (Figma’s advocates)
🔎 10 meetings > 100 (focus wins)
📈 Raw traction beats polish (Notion’s buzz)
❤️ User love outpitches you (Figma’s advocates)
🔎 10 meetings > 100 (focus wins)
March 28, 2025 at 3:19 PM
VC myths are killing your raise—here’s the truth:
📈 Raw traction beats polish (Notion’s buzz)
❤️ User love outpitches you (Figma’s advocates)
🔎 10 meetings > 100 (focus wins)
📈 Raw traction beats polish (Notion’s buzz)
❤️ User love outpitches you (Figma’s advocates)
🔎 10 meetings > 100 (focus wins)
Big LLMs are overhyped—small, adaptive AIs are the real play:
🧬 Evolves mid-task (xAI’s tech)
🎥 Redefines creative (Runway’s films)
🎉 Quietly wins (no fanfare)
🧬 Evolves mid-task (xAI’s tech)
🎥 Redefines creative (Runway’s films)
🎉 Quietly wins (no fanfare)
March 27, 2025 at 3:19 PM
Big LLMs are overhyped—small, adaptive AIs are the real play:
🧬 Evolves mid-task (xAI’s tech)
🎥 Redefines creative (Runway’s films)
🎉 Quietly wins (no fanfare)
🧬 Evolves mid-task (xAI’s tech)
🎥 Redefines creative (Runway’s films)
🎉 Quietly wins (no fanfare)
AI’s not a tool—it’s a boss. Jasper’s $100M ARR in 2024 proves it:
✍️ Writes ads solo
📈 Funds its growth
🛞 Humans just steer
✍️ Writes ads solo
📈 Funds its growth
🛞 Humans just steer
March 26, 2025 at 3:19 PM
AI’s not a tool—it’s a boss. Jasper’s $100M ARR in 2024 proves it:
✍️ Writes ads solo
📈 Funds its growth
🛞 Humans just steer
✍️ Writes ads solo
📈 Funds its growth
🛞 Humans just steer
RL’s stealing the AI show—ArXiv says it cuts costs 40% vs. LLMs. Why it’s huge:
Adapts mid-task (Tesla’s Optimus)
Outsmarts static models
Wins where Big Tech lags
Adapts mid-task (Tesla’s Optimus)
Outsmarts static models
Wins where Big Tech lags
March 25, 2025 at 3:19 PM
RL’s stealing the AI show—ArXiv says it cuts costs 40% vs. LLMs. Why it’s huge:
Adapts mid-task (Tesla’s Optimus)
Outsmarts static models
Wins where Big Tech lags
Adapts mid-task (Tesla’s Optimus)
Outsmarts static models
Wins where Big Tech lags
VCs are pickier than ever—PitchBook says Series A raises now take 5-7 months. Stop the grind:
👓 Focus beats volume
🪝 Lean pitches hook fast
👯 Network trumps cold emails
👓 Focus beats volume
🪝 Lean pitches hook fast
👯 Network trumps cold emails
March 24, 2025 at 3:19 PM
VCs are pickier than ever—PitchBook says Series A raises now take 5-7 months. Stop the grind:
👓 Focus beats volume
🪝 Lean pitches hook fast
👯 Network trumps cold emails
👓 Focus beats volume
🪝 Lean pitches hook fast
👯 Network trumps cold emails
VCs are done blindly chasing growth—First Round says 70% now prioritize capital efficiency. Here’s why it's huge:
💰 Revenue > hype
🥇 Lean teams win
🤑 Profit signals trust
The hidden truth? Relying on more funding adds risk. Capital independence is the quiet edge. Not sexy, but it works.
💰 Revenue > hype
🥇 Lean teams win
🤑 Profit signals trust
The hidden truth? Relying on more funding adds risk. Capital independence is the quiet edge. Not sexy, but it works.
March 21, 2025 at 3:19 PM
VCs are done blindly chasing growth—First Round says 70% now prioritize capital efficiency. Here’s why it's huge:
💰 Revenue > hype
🥇 Lean teams win
🤑 Profit signals trust
The hidden truth? Relying on more funding adds risk. Capital independence is the quiet edge. Not sexy, but it works.
💰 Revenue > hype
🥇 Lean teams win
🤑 Profit signals trust
The hidden truth? Relying on more funding adds risk. Capital independence is the quiet edge. Not sexy, but it works.
Big LLMs are over—ArXiv says micro-models like Mistral’s Mixtral 8x7B beat them on niche tasks with 10x less compute. Why it matters:
March 20, 2025 at 3:19 PM
Big LLMs are over—ArXiv says micro-models like Mistral’s Mixtral 8x7B beat them on niche tasks with 10x less compute. Why it matters:
Reposted by Parul Singh
Him: I've been reading a lot of post-apocalyptic stories lately.
Me: Oh, you mean the newspaper?
Me: Oh, you mean the newspaper?
March 17, 2025 at 6:02 AM
Him: I've been reading a lot of post-apocalyptic stories lately.
Me: Oh, you mean the newspaper?
Me: Oh, you mean the newspaper?
AI-native companies create compounding moats: Data flywheels, model improvements, and automation loops. SaaS? Mostly UI with APIs. Big difference.
March 11, 2025 at 3:19 PM
AI-native companies create compounding moats: Data flywheels, model improvements, and automation loops. SaaS? Mostly UI with APIs. Big difference.
This is spending I would have been happy for my tax dollars to support. Local farms + schools and food banks: seemed win/win
www.politico.com/news/2025/03...
www.politico.com/news/2025/03...
USDA cancels $1B in local food purchasing for schools, food banks
States have been notified that they will not receive 2025 funding for schools to buy food from nearby farms.
www.politico.com
March 11, 2025 at 6:22 AM
This is spending I would have been happy for my tax dollars to support. Local farms + schools and food banks: seemed win/win
www.politico.com/news/2025/03...
www.politico.com/news/2025/03...
New world order:
Software = commodity?
New platforms > old ones
GTM + data = moats
Early enterprise deals = consulting contracts
Tiny teams -> profitable fast
No one wants to IPO bc public markets are fickle, risky & onerous.
Founders know VC rules & can fund each other.
Growth VC may be redundant.
Software = commodity?
New platforms > old ones
GTM + data = moats
Early enterprise deals = consulting contracts
Tiny teams -> profitable fast
No one wants to IPO bc public markets are fickle, risky & onerous.
Founders know VC rules & can fund each other.
Growth VC may be redundant.
March 10, 2025 at 8:35 PM
New world order:
Software = commodity?
New platforms > old ones
GTM + data = moats
Early enterprise deals = consulting contracts
Tiny teams -> profitable fast
No one wants to IPO bc public markets are fickle, risky & onerous.
Founders know VC rules & can fund each other.
Growth VC may be redundant.
Software = commodity?
New platforms > old ones
GTM + data = moats
Early enterprise deals = consulting contracts
Tiny teams -> profitable fast
No one wants to IPO bc public markets are fickle, risky & onerous.
Founders know VC rules & can fund each other.
Growth VC may be redundant.