utkarsh_apoorva

AI is not conscious - its neither imaginary nor friend. It's a tool, like a knife that can cut an apple for me. Except this tool can generate text. Homo-habilis discovered tools 2.4 million years ago - let's keep it going.

Not even close. I use all of them. My claw has max context (i maintain a personal-wiki of all my projects) and does the best job. The next best thing is Claude cowork. I use chat windows of claude.ai or chatgpt only when researching general topics, not specific ones.

Harness designs assume "high-agency users". This is either bad design, or arrogance, or both.

I know the problem, because we built one for low agency users, in rural India, for financial inclusion.

Usability is a problem the AI-discourse is ignoring at its own peril.

Like the concept. This is not a business - should be an open source GitHub repo maybe.

They lost me with just one microcopy - “start earning”. Huge red signal.

I disagree with practically every take here, and likely will get hate for this message.

Personally it has been a huge unlock and here are some specific use cases.

1.Context engineering for “any task”: my claw agents debate with me, and write me files to be picked up by Claude code or Claude cowork to write code/build presentations.

2. Discovering this page: my personal newsfeed that took me 10 mins to setup. Never had any newsfeed before this which could be personalised like this.

3. Strategy: similar to point 1 above. I set up a “tough-coach” skill. Now all my 1am ideas go through the wringer of why they suck, and what, if anything could be differentiated about them.

4. Many other small-ish things that are important enough for me to spend on them, but not important enough to invest my time.

I run a conversational AI startup in India working mostly in Financial inclusion space - so most of time is spent in meetings (:-/).

With openclaw, the 2-3 hours before I sleep (low energy, high on ideas), convert into pretty productive sessions. No more “this is a great idea, I will do it tomorrow”. Instead “oh this sounded great but my agent just called it shit - what if I modify it a bit, and then let’s see”.

Absolutely. Cannot find the images that are not visible though yet. Should have taken screenshots :-(

Yes, that's pretty much what it looked like. The blank spaces you see are all product screenshots.

It used to be a typical SaaS website.

Will try to see if I have any screenshots. I must.

The conversion rates (visitor to signup) went up from around 7.4% to 11.2%. Still curious to see if this holds.

He is a non-confirmist :-D http://paulgraham.com/conformism.html

It has been pretty amazing to follow your journey, Arvid. The blog you just wrote is a godsend for anyone looking to self publish. There are so many self-publishing resources not known to most of us (editing tools, proofreader, cover design, distribution).

The section on ISBN was super critical for me personally.

It is not clear what exactly will the included in the app. Do you mean there will be design templates, like Canva, or project management templates, like Trello templates.

In the former case, I'd prefer Canva. The latter is interesting, but templates did not exactly take off with Trello.

The latter. MongoDB Charts comes the closest to what I need.

Interesting idea. Will try to do this. If I can automate it, probably that's it.

You can replace "user story" with tickets, issues, or tasks - anything that goes into Jira as a part of the sprint plan.

Thankfully, most engineers I know read, edit and mark those items as done in a project / sprint management tool.

Usually Google Docs. Lately Notion.

Jira Wiki markup seems like a viable solution, even if it has a bit of learning curve.

Haha. Thanks for the help. Shared hatred is revolutionary.

I guess it will ultimately boil down to an ETL workflow and a BI tool. Most of my requirements are fairly simple so I did not want to put up with the clunkiness of a complex software.

>Metabase works to some extent

.. and that was enough actually. I would have continued with it if it did not crash on me all the time.

Hey, thanks. Exploring Seektable.

That's cool actually. Not being in the browser is not a dealbreaker. I need something simple, not clunky, preferably free or low subscription cost. Checking out Robomongo as I write this.

EDIT: neither of these can be used for BI like visualisations, can they?

This is fairly accurate. The no code revolution is a great happenstance for most techies who want to test ideas and run experiments.

Bye bye boilerplate, until we meet again.

On that note, is there a business intelligence tool for the no-code (or low code) world that just works?

I want to visualise my database without having to type queries in terminal, and without using some clunky shit like metabase.

OP here. Thanks for the upvotes. Some directions too please :-)

Active thinking is being aware of a thought and trying to dissect the thought into its constituent parts. It's a lot like "Active Listening," except you are the only participant.

My book is called - Makers Laws - Convert Thoughts to Profitable Products.

There are three sources of my book:

My learnings from two profitable products I created (and 8+ failed ones). One of these (HandyTrain) is used by biggest enterprises in India to train their field forces - over 150K MAUs. Another one (GuitarStreet) was India's most popular online guitar store. We sold it in 2018.

Buddhist Texts - Mostly the exploration of "mindfulness of the self" (incorrectly interpreted as "mindfulness of the breath").

Term Logic of Aristotle - The methods involved in the dissection of the thoughts are inspired by this.

---

Here's an excerpt:

Reality vs Imagination The hardest thing about creating products is distinguishing Reality from Imagination. We create an imaginary problem and fall in love with it. And we refuse to believe that it is not real.

As an example, read the following:

“I invented a model of the world that didn't correspond to reality and worked from that. I didn't notice my model was wrong until I tried to convince users to pay for what we'd built. Even then I took embarrassingly long to catch on. I was attached to my model of the world, and I'd spent a lot of time on the software. They had to want it!”

This hapless young founder, so attached to his “model of the world” is Paul Graham of YCombinator.

Yes, Paul Graham.

It is an excerpt from his blog - How to Get Startup Ideas

We are wired to believe that our version of the world is the real one. And it’s super hard to get rid of this.

But there is a trick.

Classify as everything you know, as imaginary.

Really.

This sounds a bit like ancient, religious texts convincing you that nothing is real. But that's not my suggestion here.

My suggestion simply is - assume that everything you know, is not real, but instead, is a theory. It is an exercise that will help you question everything.

We will later explore how to tie this up with Active Thinking and start to see things that you never would have thought existed.

Let us look at some of the most commonly held beliefs, and try to argue that they do not represent reality.

Instead of the word “imaginary”, I will use the word Theory. A theory is a plausible explanation of a pattern, and by definition, does not claim that it is real.

Here we go.

People hate it when you lie to them - Theory. What if they are in denial, and lies feed the denial.

People hate being sold to - Theory. What if they are looking for the product you are selling

Wars are bad - Theory. Wars may reorganize the human race, sometimes for the better.

And so on...

Finding the Axioms of your Thought I am not arguing that all the above statements are false.

I am arguing that anything, when seen as a theory, becomes “conditional". The moment you call something a theory, you think about exceptions - situations when that statement could be false.

Every time you do this, you discover the conditions needed for the statement to be true.

In the language of Mathematics, you discover the Axioms of the theory.---------------

Do let me know if you liked the topic and the excerpt.

I am not sure if people need this. So as an experiment, I have written the first 4 chapters only.

And I have opened it up for Pre-orders, to gauge whether this book should exist.

EDIT:

if you want to pre-order, use the discount code "showhn" (without the quotes) - created specifically for this community ️

It seems our biological neural networks, probably, aren't all that different from the ones we write. In their 'scope' they basically do just enough, for us to function. Beyond that, evolution probably started to optimise other things.

"Just enough" would mean being able to agree upon whether the object being seen was a danger (say a wild animal, an enemy tribe) or not (a friendly tribe). This goes deep enough for us to agree upon facial expressions and the like.

Interestingly, the 'lack of agreement' would theoretically be impossible to detect, because for consensus, we would achieve "eventual consistency" of conclusions - thereby reforming the neural pathways of all individuals.

Great effort. I like the snippets from different developers in-context, instead of showing the entire interview in one place.

Why did you quote Zig Zigler? Seems a little out of place. Also, I felt that the first chapter needs to be more compelling - but maybe that's just me.

Wanting to get paid for a book he wrote is a bad thing? I get it - from each according to their capability - to each according to their needs. Sound familiar?

The lack of email walls was so refreshing. I expected an ugly form to show up, and was willing to fill it out. But instead, I was greeted with the actual download - just like that.

I don't know what the opposite of Cognitive Dissonance is called, but now I know what it feels like.

This is a pretty interesting place to be. Since you have not yet committed yourself - so you sort of have a superpower that many of us crave for.

There are at least a few ways of thinking about this.

The well documented ones 1. Think about problems you've faced. Solve it. 2. Research online, find problems, fix them (what you seem to be doing).

Less Documented ones: 1. Pick a product you like - as a customer / user. Deep dive. See what they got right and what they didn't. Who they serve well, and who isn't happy. Then build something similar in a 'different' market (what Notion did to Evernote and Roam is doing to Notion, Freshdesk to Zendesk).

2. Think of 3-5 skills you've got as an individual. Forget products for a moment. Think about people - who can you help with a combination of those skills, better than most (or cheaper than most). Talk to these people. Offer your service. Later build a product around this (Close.com, and if I remember correctly, even Kissmetrics.)

I was planning to write something along the same lines, if I make it. I guess the reason of this happening is this - after you fail, you are too busy picking up the pieces, and trying to survive.

I have written so much about small wins at my previous startup. And when it shut down, I was just trying to avoid falling.

In the current one, we built and failed with 5 products. And then got more traction in 10 days than we'd ever had ! When I put it this way, it sound like the story of success (in the making).

But it isn't. It is (if I write it), the story of how we got it wrong, and what not to do. The bonus is that we make it.

--

Check out LightCat.io to see what got traction eventually.

You guys are smart - yes, this is a shameless plug. I am poor and cheap, and this is all I've got.

If I am getting the drift right here, the root of the problem is that the Product Org is not prioritising well, and putting an upper limit on the total story points pushed in a release ?

That would lead to a large number of stories being classified as "Priority" and left to engineering to figure out how to deliver.

Separate Product Orgs, like or hate it, are not going away anytime soon. If anything, I believe they will become more ubiquitous.

Probably what we need is not more and better ticketing system, but better prioritisation systems, with well defined upper limits on how many story points can be released at once.

> I'm firmly convinced the MVP culture of launching as soon as you've built a UI with Bootstrap and thrown it up on Heroku is bonkers

So totally agree with this. All of your growth experiments fail if your product doesn't have the basic stuff right.

It works well though when you are building a tool that lets you do something you cannot otherwise - if that works, even a minimal UI release can get you users. Eg. Stripe (when it released).