AI and creativity
When a machine helps draft or design, where does authorship live? In the prompt, the selection, the editing, the lived experience—or all four?
Read the essay →Human Intelligence in the Age of AI
The machine can generate. The human still has to mean it.
An independent field notebook on artificial intelligence, creativity, learning, research and the future of work—written from the perspective of a business-school graduate learning to build in public.
Explore the ideas ↓See the experimentsNo magic prompts. No claims of superhuman intelligence. Just curious experiments, useful tools, and a human being who occasionally breaks his own website.
Start here
Five questions worth asking
AI for learning
A tutor that never gets tired can be useful. But fluency is not mastery, and an elegant answer is not necessarily a correct one.
Read the guide →Human–AI collaboration
Good collaboration is not surrendering judgment. It is knowing what to delegate, what to verify and what to own.
Read the framework →The future of work
Tasks may change faster than job titles. Human capital is more than output per hour: it includes agency, adaptability and dignity.
Read the perspective →AI and research
Use AI to challenge hypotheses and map questions—not to manufacture evidence or outsource scientific accountability.
Read the method →A living lab
What I am actually experimenting with
Independent web publishing
Building small, focused websites and learning HTML, deployment, navigation, metadata and search indexing through practice. Publishing a page is not the same as earning an audience.
Explore the main site ↗Human behaviour research
Exploring the Identity–Action Gap: why understanding what to do does not automatically translate into doing it. AI can assist thinking; behaviour still happens in the world.
Visit the behaviour lab ↗Writing and editorial craft
Using AI for outlines, counterarguments and revision while keeping the author's point of view, factual responsibility and final editorial decisions human.
Read more about my work ↗Fragrance and expression
Connecting sensory craft with digital storytelling. AI can help organize ideas about scent; it cannot smell a formula or replace real-world testing.
Explore scent ↗Operating principles
Human in the loop, not out of the picture
My working position: AI can reduce the cost of trying an idea. It does not remove the cost of being wrong.
Verify the consequential
Check dates, quotes, citations, code, medical or legal claims and anything another person might rely on.
Credit the collaboration
Be clear about AI assistance where disclosure matters. Do not imply a model independently witnessed, researched or tested something it did not.
Protect privacy
Do not paste someone else's confidential information into a tool without appropriate permission and safeguards.
Keep the human question
Faster output is not automatically a better life. Ask whether the work improves understanding, relationships, craft or autonomy.
The person behind the prompts
A commerce monkey meets a silicon monkey. 🐒
I'm Sahil Tayal, based in New Delhi. I studied commerce at Delhi University and earned an MBA from JBIMS Mumbai. My interests span organisational behaviour, human capital, writing, independent digital publishing and perfumery.
I don't come from a traditional software-engineering background. That makes these experiments interesting to me: modern tools make it easier to attempt projects, but the responsibility to learn, test and improve remains mine.
This site is a personal perspective, not an institutional research centre or an AI product vendor. The essays distinguish my interpretations from established evidence.
More about Sahil ↗Questions people ask
Will AI replace human creativity?
AI can automate parts of creative production and change creative jobs. Whether it replaces a particular role depends on tasks, economics, institutions and how people value human-made work.
Can non-programmers build websites with AI?
Yes, AI can help draft HTML, CSS and JavaScript. You still need to test the result, understand deployment and security basics, and maintain the site.
Is AI output always reliable?
No. Generative systems can produce confident but incorrect claims, fabricated references and flawed code. Verification should scale with the stakes.
Does AI make learning easier?
It can make explanations and feedback more accessible. Learning still benefits from active recall, independent problem-solving and checking your understanding without assistance.