Journey: Building a ChatGPT Competitor

nathantheboss

Registered Member
Joined
Feb 20, 2019
Messages
51
Reaction score
69
According to https://arxiv.org/pdf/2506.08872v1, when relying on LLMs like ChatGPT, the neural activity decreases.

Researchers call this phenomenon "cognitive debt."

This is the price of delegating your thinking to an LLM.

To fix this problem, I have locked myself in my bedroom for the last 6 months to build boterview, a competitor to ChatGPT that makes you smarter instead of dumber.

My target: hitting $100 million monthly revenue, if possible by the end of the year.

Currently, I am at zero dollars.

However, I have built the MVP and already at 2k users per month.

I will start marketing soon: TikTok, Instagram, SEO, ads. But to reach my goal, I traditional marketing is not enough and I need to iterate on the product and get a K-factor large enough to grow the userbase.

I will keep you updated.

Wish my luck, full speed :cool:
 
Interesting idea. The concept of building a tool that encourages thinking instead of replacing it could definitely stand out if executed well. Getting to 2k users with just an MVP is already a solid start. The real challenge will be product differentiation and strong distribution. Curious to see how it evolves. Good luck.
 
Hi) interesting concept, but there're questions... If the problem is delegating thinking to ai, then any ai tool creates that problem... If your product actually makes people smarter, what's the actual mechanic? Anyway, wishing you luck - your hustle is real!
 
Hi) interesting concept, but there're questions... If the problem is delegating thinking to ai, then any ai tool creates that problem... If your product actually makes people smarter, what's the actual mechanic? Anyway, wishing you luck - your hustle is real!
The mechanic is simple: instead of giving you the answer, it is active learning.

Think Duolingo, but the courses are created by AI.

Currently, I have implemented the following challenges: MCQs, reading comprehension test, informational, and Socratic voice to voice conversations
 
The mechanic is simple: instead of giving you the answer, it is active learning.

Think Duolingo, but the courses are created by AI.

Currently, I have implemented the following challenges: MCQs, reading comprehension test, informational, and Socratic voice to voice conversations
Got it. so it's basically AI as a tutor and not as an answer machine.
Socratic voice convos sound promising - that's really good for learning. Curious how you'll balance automation vs personalization at scale though.
And are you building any behavioral hooks to keep users coming back?
 
Got it. so it's basically AI as a tutor and not as an answer machine.
Socratic voice convos sound promising - that's really good for learning. Curious how you'll balance automation vs personalization at scale though.
And are you building any behavioral hooks to keep users coming
Thank you for the interest.

I don't know how to describe it. You can see how the project is doing here. Am open for feedback.

I have a quest system + email / notification in mind, but it is not implemented yet.
 
Thank you for the interest.

I don't know how to describe it. You can see how the project is doing here. Am open for feedback.

I have a quest system + email / notification in mind, but it is not implemented yet.
Thanks for the link)
Where do you get your information from? Is it some kind of online encyclopedias?
 
is this your own LLM or a wrapper?
It is based on claude, but I have built custom AI agent networks and custom AI tools that are connected to it.

So if you go to claude, the results will be very different from my AI
 
LLM can search web via API now, not my system

The mechanic is simple: instead of giving you the answer, it is active learning.

Think Duolingo, but the courses are created by AI.

Currently, I have implemented the following challenges: MCQs, reading comprehension test, informational, and Socratic voice to voice conversations
You mentioned Duolingo - I think gamification like Duolingo is a great idea. They have points and stuff , it gets you hooked.
 
you got 2k users because of forcing to login / register.
but how many people really active on the website after first use?
 
Okay but competing with the big giants? Nah I wouldn't advice that. But looks like a good concept though but hope you have lots of training infra
 
According to https://arxiv.org/pdf/2506.08872v1, when relying on LLMs like ChatGPT, the neural activity decreases.

Researchers call this phenomenon "cognitive debt."

This is the price of delegating your thinking to an LLM.

To fix this problem, I have locked myself in my bedroom for the last 6 months to build boterview, a competitor to ChatGPT that makes you smarter instead of dumber.

My target: hitting $100 million monthly revenue, if possible by the end of the year.

Currently, I am at zero dollars.

However, I have built the MVP and already at 2k users per month.

I will start marketing soon: TikTok, Instagram, SEO, ads. But to reach my goal, I traditional marketing is not enough and I need to iterate on the product and get a K-factor large enough to grow the userbase.

I will keep you updated.

Wish my luck, full speed :cool:
First of all, best of luck to you...

But my advice is to just concentrate your work. Never focus on beating ChatGPT or getting 100 million monthly revenue, just focus on what you can do best and improve your knowledge.

I appreciate your thinking and challenging guts...

Once again best of luck for your journey
Okay but competing with the big giants? Nah I wouldn't advice that. But looks like a good concept though but hope you have lots of training infra
Yeah,

That's what I say because it's not easy to beat a big giant. Because ChatGPT isn't alone in the market, there are more competitors like it
 
Back
Top