2025-11-28 08:00:00
Since I started integrating AI deeply into my workflow, my baseline for quality has shifted dramatically.
A few days ago I was iterating on a single UI card. I kept taking screenshots, dropping them into NotebookLM, feeding it references (books, patterns, screenshots), and asking, “What’s missing? What would a good designer point out?” I did the same loop across a few other tools too. My goal was to reach 70-80% of a professional designer’s output.
Most models caught things I missed. And the more context I gave, the sharper the responses.
Ryo from Cursor recently wrote that in the future, teams will be smaller, with more generalists.
If that is true, then what happens to the average specialist?
If I can generate a better artefact with just a few prompts than what a designer creates, I’m going to be disappointed with that person’s output.
The baseline is higher than it has ever been.
AI has fundamentally broken the “average” tier of creative work. Average design and average landing pages don’t hit the same, because “average” is something a model can do by default.
So, where is the alpha?
I recently saw Alter Magazine’s website. The hero section. The typography, all the imagery. The website just had soul.
It felt like it was created by someone who really cared. Yes, AI can generate. It can remix. It can get you to “pretty good” fast. But the real alpha is going beyond what the average model would produce: making choices that are surprising, but clearly intentional.
That is what models can’t replicate yet. The alpha is in taste, curiosity, and the ability to craft something that feels human in a sea of AI content.
2025-11-20 08:00:00
“Oh Manas, why do you keep posting these screenshot essays on twitter? Just write a normal long post, it is better for distribution.”
No, you do not understand. I am a Figma Pro user. Like every middle class Indian who has paid for a tool, I must extract full value from my subscription.
That is why:
Thank you for your attention to this matter.
2025-11-19 08:00:00
Don’t use NotebookLM only to consume others’ ideas. Use it to clarify and compress your own thinking so that you can present it better.
A lot of people use NotebookLM only for consumption. Drop in a paper, get a summary, maybe a video or slides.
But where NotebookLM really shines is creation.
A friend showed me his workflow around using NotebookLM to produce the first version of anything he needs to explain. He’ll drop in his notes, drafts, and prototypes, and NotebookLM gives him a tight five mins video brief that he refines to come up with the final output that he presents to his CXO.
I have started doing the same. I had a long manifesto for my startup and wanted to turn it into a short elevator pitch. NotebookLM compressed it perfectly and gave me a clean base to iterate from.
2025-11-18 08:00:00
The fastest way to get started with a new project, especially if you want to copy an existing product and iterate on top of it:
Use the Magicpath Chrome extension With this extension, you can select an element on a webpage and copy-paste it into Magicpath. It will automatically generate a React component that you can use. However, there’s a limitation: while you can view the component, you can’t export the code unless you pay for Magicpath. The output is pretty good. Trust me, I’ve tried many vibecoding tools. It’s still hard for them to visit a website and copy either the whole page or the individual components. And yes, I know Cursor can now see your browser, it still doesn’t work.
Figma Make -> Figma Design -> Claude Code Another approach is to take screenshots of the components you want. Then, go to FigmaMake and ask it to create the component from the screenshot. You can one click export the component from Figma Make to Figma Design. If you’re a designer who prefers working in Figma instead of dealing with code, this method gives you more control. You can then refine the component directly within the Figma design environment. Once you’ve finalised the design and are ready to start building, you can use the Figma MCP + Claude Code client to build a working prototype.
2025-11-17 08:00:00
It is not typography. It is not colors. It is not your graphic design blowing up on Twitter.
It is who you are. And it is how your users feel at every point of interaction. Your landing page. Using your product. Reading your Twitter.
Take prediction markets. Should a prediction market feel like a game or should it feel like a Bloomberg terminal?
Depends.
If you see prediction markets as a comprehensive venue where all assets get traded but for now you see it as an extension of your core like Robinhood does, then it will be just an extension of your existing product and brand.
If you see prediction markets not as a financial primitive but as a real money gaming product where people come to speculate and have fun, then it can look like a game.
You can go as crazy as making the interface look like Persona 5.
Yes, it will turn some people off but you will be unmistakably you. N of 1.
But if you want to be seen as the source of truth and people are going to take policy decisions based on your odds, and your cards are going to be attached as widgets on publications like Wall Street Journal, then making a Persona 5 vibes widget (that is an extension of your product) might not reflect well.
It is all the small decisions. It is not “thinking”, but “frolicking” in Claude Code.
Claude wasting a ton of space on homepage to say “Welcome back, night owl” when it probably does not affect their product metrics.
Cursor running their pop up cafes to attract their fans.
The same thinking applies to people too.
Everything you post online. Every interaction you have offline. Is a reflection of who you are. It is not what you wear. It is not how well you speak.
It is how you make the other person feel.
Do I post salty takes sometimes? Yes. Can I not post those? Sure. But then I won’t be me. And life is too short to be not yourself.
“Be yourself” could actually be destroying your optionality.
Cluely had a strong consumer brand, love it or hate it, everyone knew what it stood for. But you can’t be Cluely and then suddenly try to become an enterprise B2B company.
The same applies to Marc Benioff, who faced backlash when he attempted to change his political stance. Or Sam Altman, who was criticised for asking Elon’s partner if he should post something nice about Elon. In today’s world, people spot hypocrisy quickly.
And here’s another question to consider:
Does what I believe to be my authentic self-expression truly serve the business objective, or does it simply satisfy my personal preference and ego?
Maybe it can be both. By being unabashedly me, I can play long term games with long term people. It reduces my optionality. But I don’t need to be another grin fucker :)
2025-10-11 08:00:00
I have tweeted about this many times in the past. This is such a big opportunity that no one has nailed so far here.
Palantir won’t bother servicing companies in Southeast Asia. They remain focused on the West, mainly for cultural reasons and because of the larger deal sizes.
OpenAI, on the other hand, won’t send a full deployment engineering team unless the deal size is around $10 million (if I’m not mistaken). Most AI curious companies are just using these LLMs to chat, but have not got the most out of them.
People talk about building “Infosys 2.0,” but doing that requires more than just a team of smart techbros, you also need a founding team with the right connections to open doors. (Take a look at the composition of the Brainco team for example from the transcript below.)
I heard rumors that a VC firm was seriously considering bringing together a group of AI experts from Silicon Valley to launch something similar in India. But people are underestimating how difficult it is to sell large-scale transformation projects.
You need both capability and access.
Infosys and the other WITCH companies can’t pull this off. They simply don’t have the muscle for it. This will have to come from a completely different company altogether.
“TBPN: You’ve been working with Jared Kushner on a new company. I know he’s been busy, but I’d love to hear the latest.
Elad Gil: Sure. We recently launched a company called Brainco, which focuses on using AI as a platform to help transform the world’s largest institutions. We’ve been working with a number of large enterprises, private equity firms, and others. It’s been a fun project to build alongside Jared, Eric Wu, and Luis Videgaray. Eric was the former CEO of Opendoor, and Luis served as the Finance and Foreign Minister of Mexico. It’s an interesting mix of people coming together to solve some of the biggest AI challenges.
TBPN: So was this born out of realizing how much firms like Accenture or McKinsey were charging to make “AI strategy” pitch decks—and thinking you could do it better by actually building software for these organizations?
Elad Gil: It’s definitely more focused on the software side. We’ve built a common platform that helps enterprises manage different forms of data, evaluation frameworks, and the other core systems they need to truly adopt AI. On top of that, we build vertical-specific applications—and sometimes horizontal ones—that can be reused across similar companies in the same industry.
For example, in financial services, there are a dozen things every major firm needs to build. There’s always some customization, but the core infrastructure remains the same. Think about large enterprise deals—if you’re Dell, VMware, or Oracle, and you’re doing a $50 million contract, you’ll absolutely tailor parts of it to the client. We’re similar in that way: a shared platform and vertical solutions, with client-specific customization where it matters.
TBPN: How are you thinking about your target customers? Are you focusing on certain verticals, like avoiding healthcare because of HIPAA, or defense because of FedRAMP? Or are you segmenting more by company size—say, only working with large enterprises, not mid-market firms?
Elad Gil: The goal is to work primarily with the world’s largest institutions—those with significant revenue, market cap, or impact. There will always be exceptions, but we’re focused on helping organizations that can use AI to create real leverage at massive scale.
TBPN: That makes sense. There was a report that JPMorgan is spending about $2 billion a year on AI to save roughly the same amount through automation.
Elad Gil: Exactly. They’re basically breaking even—though some of those savings are likely durable—but it shows how much value and investment is already being driven by enterprise AI adoption. (Note: Transcript cleaned up using ChatGPT)”