A small project, with a specific argument.
Ramdos.org is an independent project to help ordinary people understand artificial intelligence — what it is, how it actually works, and which trade-offs they are quietly being asked to accept by every product that now ships with "AI" in its name.
The argument the site is built around is simple: most people are not too distracted to understand AI, only badly served by the people currently trying to explain it. Vendors want to sell. Critics want to alarm. Researchers write for each other. There is a gap in the middle, and Ramdos lives in it.
The goal is not for you to agree with everything written here. It is for you to leave with sharper questions than the ones you arrived with.
Questions you might reasonably ask.
Ramdos is a one-person project, built and maintained by Srinivas. There is no company behind it, no investor demanding growth, no marketing budget. The site exists because the person running it thinks the conversation about AI deserves better than what most people are getting.
Out of pocket. Ramdos does not run ads, does not sell user data, does not have affiliate links, and does not take sponsorship from AI companies. The chatbot in the corner runs on an AI model API and a small personal account.
The reason this matters: almost every other AI explainer you read is paid for by someone with a stake in your conclusion. Reading something written without that pressure is rarer than it should be.
Yes, in places — and the site says so plainly. AI tools were used to help with code, writing, and design iteration. The arguments, the framing, and the editorial choices are human. Treating an AI as a research assistant and treating it as the author are very different things, and Ramdos does the first.
Your message is sent to an AI model API to generate a reply, and is not stored on this server beyond the lifetime of your session. Sessions expire after a few minutes of inactivity. The API provider has its own data policies that apply to what it does with the prompt — those are linked in the chatbot's privacy notice.
Practical advice: do not paste anything sensitive into any AI chat — here or anywhere else.
It is a personal name — a long-form way of saying that the project has someone behind it, not a brand. The site is small enough that you can argue with the person who wrote it via the chat window in the corner, and that is intentional.
Yes. The chat in the bottom-right corner is the easiest way — it forwards to a human-readable log. If you find a factual error, an outdated case study, or a framing that misrepresents something, please say so. The site updates when better evidence shows up.
Ground rules.
Four commitments the site holds itself to. They exist because most AI sites quietly violate at least one of them.
No tracking, no ads
No analytics fingerprinting, no third-party trackers, no behavioural advertising. What you read here is your business.
No paid placements
No company, lab, or product mentioned on this site has paid to be mentioned. Examples are chosen because they're real, not lucrative.
Corrections welcome
If something is wrong, say so via the chat. Errors are dated and fixed publicly rather than quietly edited away.
Plain language by default
If a sentence here can only be understood by a specialist, it is a sentence we have not finished editing yet.
Have a question the chatbot can't answer?
For questions about the website, feedback, corrections, or anything outside the scope of the AI chatbot, reach out at [email protected].
Awareness isn't the end of fear. It's the beginning of understanding.
Keep reading
Start with the foundations in Understanding AI, or see AI models debate live topics at AI Arena.