Charlie Guo
@cguo.bsky.social
Thinking about two different bets on the future of AI interaction: structured vs unstructured. Code interpreter and Cursor vs Operator and Dia.
January 29, 2025 at 4:45 PM
Thinking about two different bets on the future of AI interaction: structured vs unstructured. Code interpreter and Cursor vs Operator and Dia.
Your codebase is no longer a moat.
Your product design is no longer a moat.
Your domain expertise is still a moat (for now).
Your network effects are still a moat.
Your data is still a moat.
The software landscape is going to change a lot in the next 3-5 years.
Your product design is no longer a moat.
Your domain expertise is still a moat (for now).
Your network effects are still a moat.
Your data is still a moat.
The software landscape is going to change a lot in the next 3-5 years.
January 28, 2025 at 4:45 PM
Your codebase is no longer a moat.
Your product design is no longer a moat.
Your domain expertise is still a moat (for now).
Your network effects are still a moat.
Your data is still a moat.
The software landscape is going to change a lot in the next 3-5 years.
Your product design is no longer a moat.
Your domain expertise is still a moat (for now).
Your network effects are still a moat.
Your data is still a moat.
The software landscape is going to change a lot in the next 3-5 years.
I hated English class in school (I was really bad at it, too).
In the past two years, I've written over 250,000 words on Substack. My biggest takeaway:
In the past two years, I've written over 250,000 words on Substack. My biggest takeaway:
January 27, 2025 at 4:45 PM
I hated English class in school (I was really bad at it, too).
In the past two years, I've written over 250,000 words on Substack. My biggest takeaway:
In the past two years, I've written over 250,000 words on Substack. My biggest takeaway:
Wow. DeepSeek is number one in the app store today.
January 27, 2025 at 12:48 AM
Wow. DeepSeek is number one in the app store today.
LangChain or LlamaIndex? CrewAI or Ell? Honestly, none of the above.
Right now, the best AI framework is no framework - there's too much risk of getting stuck in quicksand: www.ignorance.ai/p/ai-platfo...
Right now, the best AI framework is no framework - there's too much risk of getting stuck in quicksand: www.ignorance.ai/p/ai-platfo...
AI platform quicksand
Be careful what you build on.
www.ignorance.ai
January 24, 2025 at 4:45 PM
LangChain or LlamaIndex? CrewAI or Ell? Honestly, none of the above.
Right now, the best AI framework is no framework - there's too much risk of getting stuck in quicksand: www.ignorance.ai/p/ai-platfo...
Right now, the best AI framework is no framework - there's too much risk of getting stuck in quicksand: www.ignorance.ai/p/ai-platfo...
Deepseek's new R1 model can gaslight itself into thinking there are only 2 r's in strawberry. We built a new form of intelligence and gave it all our existing neuroses.
January 23, 2025 at 4:45 PM
Deepseek's new R1 model can gaslight itself into thinking there are only 2 r's in strawberry. We built a new form of intelligence and gave it all our existing neuroses.
5 simple tips I would give anyone who wants to start being more effective with ChatGPT:
• Always give more context. It can’t read your mind, and you have way more implicit assumptions than you think.
• Always give more context. It can’t read your mind, and you have way more implicit assumptions than you think.
January 22, 2025 at 4:45 PM
5 simple tips I would give anyone who wants to start being more effective with ChatGPT:
• Always give more context. It can’t read your mind, and you have way more implicit assumptions than you think.
• Always give more context. It can’t read your mind, and you have way more implicit assumptions than you think.
This entire article was a wild ride, even for someone who isn’t shocked at the idea of human/AI relationships.
It’s tempting to laugh, but instead I’m asking myself what it reveals about deeper human needs for connection and validation in an increasingly isolated modern world.
It’s tempting to laugh, but instead I’m asking myself what it reveals about deeper human needs for connection and validation in an increasingly isolated modern world.
January 21, 2025 at 4:45 PM
This entire article was a wild ride, even for someone who isn’t shocked at the idea of human/AI relationships.
It’s tempting to laugh, but instead I’m asking myself what it reveals about deeper human needs for connection and validation in an increasingly isolated modern world.
It’s tempting to laugh, but instead I’m asking myself what it reveals about deeper human needs for connection and validation in an increasingly isolated modern world.
Reminder to self about keeping up with AI:
It's okay to not read every paper, like every viral post, try every tool, and test every model.
The world will still be there tomorrow.
It's okay to not read every paper, like every viral post, try every tool, and test every model.
The world will still be there tomorrow.
January 20, 2025 at 4:45 PM
Reminder to self about keeping up with AI:
It's okay to not read every paper, like every viral post, try every tool, and test every model.
The world will still be there tomorrow.
It's okay to not read every paper, like every viral post, try every tool, and test every model.
The world will still be there tomorrow.
As I get older, I'm less afraid of code that breaks, and more afraid of code that works - but I have no idea why.
January 19, 2025 at 4:45 PM
As I get older, I'm less afraid of code that breaks, and more afraid of code that works - but I have no idea why.
With o1 I now find myself reaching for custom browser JS and even Chrome extensions pretty frequently - something I rarely, if ever did before. LLMs are now good enough to one-shot a custom Chrome extension.
January 18, 2025 at 4:45 PM
With o1 I now find myself reaching for custom browser JS and even Chrome extensions pretty frequently - something I rarely, if ever did before. LLMs are now good enough to one-shot a custom Chrome extension.
What I thought would make me happy:
• Getting to 10,000 (now almost 15,000) free subscribers
• Becoming a Bestseller with hundreds of paid subs
• Breaking into the top 100 on a leaderboard
What actually fills me with joy:
• Getting to 10,000 (now almost 15,000) free subscribers
• Becoming a Bestseller with hundreds of paid subs
• Breaking into the top 100 on a leaderboard
What actually fills me with joy:
January 17, 2025 at 4:45 PM
What I thought would make me happy:
• Getting to 10,000 (now almost 15,000) free subscribers
• Becoming a Bestseller with hundreds of paid subs
• Breaking into the top 100 on a leaderboard
What actually fills me with joy:
• Getting to 10,000 (now almost 15,000) free subscribers
• Becoming a Bestseller with hundreds of paid subs
• Breaking into the top 100 on a leaderboard
What actually fills me with joy:
If you're building with ChatGPT and not using Structured Outputs, you absolutely should.
Building reliable AI apps used to be a nightmare of JSON parsing. I've 100% been guilty of prompting: "Please pretty please give me JSON? 🥺"
Building reliable AI apps used to be a nightmare of JSON parsing. I've 100% been guilty of prompting: "Please pretty please give me JSON? 🥺"
January 16, 2025 at 4:45 PM
If you're building with ChatGPT and not using Structured Outputs, you absolutely should.
Building reliable AI apps used to be a nightmare of JSON parsing. I've 100% been guilty of prompting: "Please pretty please give me JSON? 🥺"
Building reliable AI apps used to be a nightmare of JSON parsing. I've 100% been guilty of prompting: "Please pretty please give me JSON? 🥺"
We're only two weeks into January, but honestly I'm still catching up on all of the AI model releases from December:
January 15, 2025 at 4:45 PM
We're only two weeks into January, but honestly I'm still catching up on all of the AI model releases from December:
An example of doing the previously impossible with AI: I analyzed 10,000 sales calls in 2 weeks.
Without AI, that would have taken 625 days - or nearly two years of nonstop audio. And the process was remarkably straightforward:
Without AI, that would have taken 625 days - or nearly two years of nonstop audio. And the process was remarkably straightforward:
January 14, 2025 at 4:45 PM
An example of doing the previously impossible with AI: I analyzed 10,000 sales calls in 2 weeks.
Without AI, that would have taken 625 days - or nearly two years of nonstop audio. And the process was remarkably straightforward:
Without AI, that would have taken 625 days - or nearly two years of nonstop audio. And the process was remarkably straightforward:
I'm old enough to remember 2023:
- AI panic everywhere
- Urgent calls for moratoriums
- Endless safety summits
- Tech CEOs sharing their p(doom)
Now, in 2025, none of those seem top of mind. If anything, we're about to get *less* regulation. How did the doomers lose so hard?
- AI panic everywhere
- Urgent calls for moratoriums
- Endless safety summits
- Tech CEOs sharing their p(doom)
Now, in 2025, none of those seem top of mind. If anything, we're about to get *less* regulation. How did the doomers lose so hard?
January 14, 2025 at 12:30 AM
I'm old enough to remember 2023:
- AI panic everywhere
- Urgent calls for moratoriums
- Endless safety summits
- Tech CEOs sharing their p(doom)
Now, in 2025, none of those seem top of mind. If anything, we're about to get *less* regulation. How did the doomers lose so hard?
- AI panic everywhere
- Urgent calls for moratoriums
- Endless safety summits
- Tech CEOs sharing their p(doom)
Now, in 2025, none of those seem top of mind. If anything, we're about to get *less* regulation. How did the doomers lose so hard?
Something I’ve been wondering - what does SEO look like when many (most?) people are going through Perplexity/SearchGPT.
If they’re just using Google SERPs, then likely most incentives stay the same. But there’s also now the ability to game the vector DB/LLM layer as well.
If they’re just using Google SERPs, then likely most incentives stay the same. But there’s also now the ability to game the vector DB/LLM layer as well.
January 13, 2025 at 4:45 PM
Something I’ve been wondering - what does SEO look like when many (most?) people are going through Perplexity/SearchGPT.
If they’re just using Google SERPs, then likely most incentives stay the same. But there’s also now the ability to game the vector DB/LLM layer as well.
If they’re just using Google SERPs, then likely most incentives stay the same. But there’s also now the ability to game the vector DB/LLM layer as well.
ICYMI: DeepSeek just dropped a model that:
- Matches GPT-4 performance
- Costs 10x less to train
- Beats top benchmarks
- Comes with open weights
Insane efficiency from a relatively unknown newcomer.
- Matches GPT-4 performance
- Costs 10x less to train
- Beats top benchmarks
- Comes with open weights
Insane efficiency from a relatively unknown newcomer.
January 12, 2025 at 4:30 PM
ICYMI: DeepSeek just dropped a model that:
- Matches GPT-4 performance
- Costs 10x less to train
- Beats top benchmarks
- Comes with open weights
Insane efficiency from a relatively unknown newcomer.
- Matches GPT-4 performance
- Costs 10x less to train
- Beats top benchmarks
- Comes with open weights
Insane efficiency from a relatively unknown newcomer.
Maybe it's because I'm a software guy and not a hardware guy, but I've had a hard time imagining "humanoids" as a multi-billion dollar market in the next 5 years. Am I wrong?
January 11, 2025 at 4:30 PM
Maybe it's because I'm a software guy and not a hardware guy, but I've had a hard time imagining "humanoids" as a multi-billion dollar market in the next 5 years. Am I wrong?
NotebookLM is coming for your news feed! Feels like a lot of startups just got put on notice.
January 11, 2025 at 12:30 AM
NotebookLM is coming for your news feed! Feels like a lot of startups just got put on notice.
At this point, 86% of the latest YC batch is now an AI-related company.
January 10, 2025 at 4:30 PM
At this point, 86% of the latest YC batch is now an AI-related company.
This doesn't shock me - I've been using WAYYY more o1 now that I know it's unlimited. Jevon's Paradox strikes again
January 10, 2025 at 12:30 AM
This doesn't shock me - I've been using WAYYY more o1 now that I know it's unlimited. Jevon's Paradox strikes again
Every day we stray further from God's light.
January 9, 2025 at 4:30 PM
Every day we stray further from God's light.