Episode 131: The Rise of “Vibe Founding” with Samim Safaei

We were fortunate enough to have a bunch of founder friends or networks around us that we leveraged, so we started doing a lot of customer interviews, customer discovery, and validation work and got a range of different things.

One of the patterns we saw is most founders don't even know what they need help with and sometimes have very superficial things like needing help with fundraising, but if you're really hitting a wall, maybe there's an underlying reason such as not being ready, lacking the right network or contacts, or not having the traction you need yet.

My name is Demetri Panici and I'm a content creator, agency owner, and AI enthusiast. You're listening to the AI Agents Podcast brought to you by Jotform and featuring our CEO and founder, Aytekin Tank. This is the show where artificial intelligence meets innovation, productivity, and the tools shaping the future of work. Enjoy the show.

Hello and welcome back to another episode of the IHS podcast. In this episode, we have the founder and CEO, Sim Safay, of Sift AI, and we're really excited to have him on today. How are you doing, Sim?

Doing good, Demetri. How are you?

I'm doing awesome. Really appreciate you making the time. I know we've ping-ponged this a little bit for a while, so I'm really excited to finally get chatting with you. What would you say is your story as to how you got into AI and then how you had the idea for Sift?

Sure. I got into AI a couple startups ago. Sift is my fifth startup but my third was actually in the music space where we developed a music visualization platform like a smart disco ball that visualizes the emotion in music. This was pre-GPT days and involved training an ML model to map audio spectrum to visual spectrum.

That got my curiosity peaking because I'm an engineer by trade but very product focused and trending more towards product over time. I started to see how AI could fill gaps and solve tough problems.

Years ago, after GPT came out and all the generative AI stuff, there was a huge boom in what people could create. I was tinkering with different tools but the support and guidance entrepreneurs had was still minimal, feeling like when I was doing my first startup.

Even though startups are more mainstream now with lots of founders, there's still a lack of support and guidance. We thought about how ChatGPT and Claude are cool but not really geared towards the founder's journey or helping people turn ideas into viable businesses.

That was the genesis of Sift, started just under a year ago last May. My team and I have experience building products and being founders before the AI age, so we know the frameworks and are adapting those to a more AI native workflow and tool.

Talk a little about what it was like before and how you've seen it change and adapt. Many people only know AI where we are now and don't see the transition.

Back in the day, we followed frameworks like Lean Canvas and read books like Lean Startup, manually doing strategy and thinking about building a business. It was slow and granular, which was good because it made you think, but there were many barriers and less could get done in a day or week.

Nowadays, in our last project before this, we used ChatGPT for quick market research and brainstorming, but as you evolve your business and iterate, the AI tools aren't really built for that iterative ideation and context. They hit memory and scaling problems, so we couldn't pivot or evolve our business easily.

All our context and intel was baked into the project and couldn't be surfaced or adapted easily, so in some ways we were flying more blind using ChatGPT as founders than before having those tools.

We talked prior about ChatGPT and maybe the inevitable downfall of OpenAI, just kidding. When building a product, how did you figure out priorities for people versus what's already out there?

We were fortunate to have founder friends and networks we leveraged, doing lots of customer interviews, discovery, and validation work. We saw that most founders don't know what they need help with and have superficial issues like fundraising, but often there's an underlying reason like not being ready, lacking network, or traction.

It's tough to tell people that and articulate it, especially as founders who get precious about their ideas and find it hard to trust advice from those without context or insight.

Many founders are isolated and don't clearly understand their problems. We saw patterns like not knowing what to ask chatbots, having directional issues, obsessing over product prematurely, or doing shallow validation.

I'm guilty of hearing what I want in interviews or asking biased questions. ChatGPT can be a yes-man, preferring new ideas even if worse, showing bias and people-pleasing, which is tricky.

Vibe coding and vibe marketing are big now, and vibe revenue ties into vibe founding, which is building a business based on vibes, emotions, and hunches rather than logic and data. I'm an OG vibe founder but it's tricky now because chatbots tell you everything you're doing is right, amplifying biased feedback.

Vibe founding is amplified by AI chatbots, creating echo chambers and amplified noise. I agree there's a feedback loop where AI talks to itself repeatedly.

I do content and SEO, and many tools use sources like Reddit because Reddit has an anti-AI layer, unlike blogs generated by AI which are less trusted now. Using web search with LLMs initially added external context but then retweaked to add Reddit, causing repetitive info.

Reddit info can be false frequently and is colloquially joked about as 'Reddit tier arguments.' It's a funny world where the system is self-affirming if unchecked.

Echo chambers are getting more robust and noise amplified.

What are some cool features you're bringing and who are the main users or industries for your product?

Vibe founders are a natural user group, especially non-technical or first-time founders empowered to start projects and businesses but lacking experience, training, or time to think about strategy, go-to-market, or validation.

Solopreneurs are another big group who struggle to filter biases and blind spots and lack people to walk through things with except yes-man chatbots. Our platform helps filter against that and think through strategy logically and granularly.

We offer structure as a service with structured databases, tables, and unstructured chat data. Sometimes you need structure in unstructured chats. ChatGPT business plans have valuable nuggets but also lots of noise that needs updating.

Our platform sifts and manages data, keeps it updated and validated. The UI is proprietary, creating a hybrid between chat and structured tools like Notion or ASA.

Many AI tools try to balance being a great theoretical tool with many connections and doing everything by chat, but also having structure.

Chat is essential, maybe voice in the future, but pictures or visual displays like canvases, tables, or dashboards are more efficient and intuitive than scrolling through conversation threads.

Chat is the main communication mechanism but there's a filtering or extraction layer pulling info into structured views or dashboards to maintain big picture awareness.

Founders have parallel ideas and don't know which is right, with assumptions that may both sound correct but differ in importance. Managing parallel context threads and updating context as validated or disproven over time is key.

Traditional chatbots have one plane of truth and context, so pivoting causes contradictions and hallucinations. Our weighted context system can weight importance, relevance, and validity as thinking and business evolve.

This system is something we're trying to patent to protect our mission and enterprise aimed at empowering founders.

Many AI tools have scope creep and don't get users anywhere. How will you make it easy for founders to be empowered and avoid these pitfalls?

We call it the Zork problem, named after a 1970s text-based video game with minimal info forcing creative thinking. Founders often don't know what to ask or ask wrong things, staring at a blank conversation line like Zork.

The best way to empower founders is by giving them the right questions or suggestions to ask, helping them start going in the right direction.

ChatGPT and Claude are general purpose tools with directionality or agency problems, but Sift is specialized to help derisk, accelerate, and succeed in the founder's journey with playbooks, frameworks, and step-by-step guidance.

Founders have unique needs compared to general work or productivity apps, making specialized systems like Sift advantageous.

The work founders do is specialized because building something specialized is required or the business is dead on arrival.

Sift started last May and is under a year old. Do you use it yourself and what does that look like?

We're starting to use it internally. There have been delays due to backend complexity, but as the tool becomes available, we're seeing granular thinking and dependency mapping not found in chatbots.

For example, if I want to test TikTok as a channel, ChatGPT or Claude will give plans, but Sift will question if TikTok fits the audience, helping refine strategy and coherence with deeper ontology and dependency maps.

What is your goal for the final or current product framework?

It will be more guided and automated, integrating crowdsourced automations and skills seen in ecosystems like Claude.

Our vision is for Sift to be the founders' platform, not specializing in market research or vibe coding but as an intelligent filter to kickstart and derisk projects.

Eventually, we want to integrate useful third-party tools and ecosystems to empower rather than compete with communities like vibe coding and solopreneurs.

How do you think the solopreneur landscape will change in 2026 with AI agents being big in 2025?

The solopreneur market has boomed with vibe coding and AI tools breaking barriers to enable more people to build businesses and be their own boss.

The hype and froth of 2025 will melt in 2026 as many tools are similar and undifferentiated, leading to commodification and niche or strategic offerings.

The market will mature and continue growing but likely plateau rather than explode exponentially, possibly following an S-curve.

Tools like Sift solve problems for solopreneurs and small businesses, changing the landscape. The future is entrepreneurial and small business focused.

Large organizations' pilot failures are often people, adoption, and implementation problems, classic issues not technology problems.

As enterprises adopt AI, more small businesses will succeed, increasing the competitiveness gap between small businesses and enterprises, with legacy enterprise issues worsening.

Do you think enterprises will struggle to keep up with lean teams?

If a one-person team does the work of ten and earns revenue like a company of 100 or 200, how do enterprises compete, especially if bootstrapped and without financial pressures?

Layoffs and other enterprise issues create negative feedback loops. Job security and perks of big companies are eroding as AI adoption grows into 2026.

Working at enterprise often means moving from operational roles to middle management, where many lose skills over time.

Middle management is colloquially known as cushy because they often do little output, making it easy for companies to cut costs by replacing them with cheaper resources.

Layoffs in operational roles are more upsetting, but middle management often lacks fulfillment and passion for their jobs.

People dislike jobs more due to lack of fulfillment than pay. Middle management often lacks passion, and layoffs might be a wake-up call to improve skills.

Layoffs often target middle management, which some see as a positive push for skill improvement.

People moving from companies like Microsoft to project management roles sometimes lack recent technical skills, showing the need for skill updates.

Layoffs push people to be better, and relationships with work and careers are shifting due to new tools and possibilities.

Job security is a false pretense; many are only a few quarters away from layoffs, so it's better to focus on what you want to do and leverage AI to do it.

Having a few thousand subscribers or customers can replace income from traditional jobs, showing the potential of niche markets.

You've been working on Sift yourself for under a year. Did you previously work in a more secure position?

Yes, I worked at a large marketing agency at the operator level, which was more secure. I appreciate taking the leap to entrepreneurship.

A big mental model was the 'what's the worst that could happen' approach, with savings prepared for a year or two.

The company I worked at had a massive layoff after being bought by private equity, showing that job security is a farce unless you run your own company.

Job security is outdated and not viable anymore due to restructuring, M&A, and massive changes in organizations.

Organizations are becoming more bottom-heavy with AI agents replacing many roles, forcing people to rethink job security.

The best job security is working for yourself.

If there was one thing to recommend before checking out your product, what would it be and where should people go?

Ask yourself what you want in a founder platform and how you can trust it to be founder-first and keep your best interests at heart.

I've spent over a decade doing startups, and Sift aims to actively change and improve issues in entrepreneurship, empowering people to build businesses rather than just scaling for profit.

More information is on the website, and people can sign up for early access with the code AI Agents for a discount and promo from this episode.

Make sure to check out Sift AI spelled S-I-F-T, and please like, subscribe, and leave reviews on Apple Podcasts and Spotify. We appreciate the work you're doing, Sim, and thanks to everyone for listening. See you next time.