Michael Fallon

Building @korvaapp

[Michael Fallon @_michaelfallon]

I’m thinking about changing how I market Korva and how I talk about it. I got too focused on explaining what it does, rather than the value people get from it and how I want Korva to make them feel.

I’ve made adding a podcast extremely easy. But pitching it as a way to “save the insights from your podcast” still sounds like work. It makes adding something feel like the starting point.

My grand vision for Korva has always been a place with the internet’s best insights, without the slop.

We all know and love the experience of coming across something interesting online. A juicy detail, a fun fact, or an idea that makes you curious and leaves you wanting to know more. I want Korva to give people more of that, with less slop between the good stuff.

Open it and scroll. The insights are one or two sentences. You can go deeper, return to the original source, and discuss them with other people as the community grows.

For some people, it can be a companion to what they already watch, read, or listen to. For others who are too busy to listen, read, or stay up to date, it can still be a way to find something fun and interesting, and give that curious side of you something to enjoy. And if you’re bombarded with things to consume, it can help you decide what deserves your time and discover something you become a fan of.

We’ll see whether this gets more people interested in Korva, but I think it captures the value better than leading with the utility.

So my question: what would make you more interested? An app positioned as a way to save insights from podcasts, or a collection of the internet’s best insights, where you can get more of those moments we all love: coming across an interesting tidbit, a fun fact, or an idea you want to know more about?

[Michael Fallon @_michaelfallon]

The related Korva insights:
Range anxiety: https://t.co/IcBlejLNh1
Choosing trade-offs: https://t.co/NufomVsMsW

Full conversation: https://t.co/zkFNJr8bCy

[Michael Fallon @_michaelfallon]

User expectations matter, and so does the way a product helps people understand what they’re getting. On My First Million, @rorysutherland gives @theSamParr and @ShaanVP examples involving electric cars, coffee and hotels.

He describes getting nervous when his Lotus dropped to 16% battery. The display had turned amber, and he was starting to worry. Then he checked the estimated range and saw about 58 miles left.

He says he and his wife regularly drove her Mini with about 56 miles left without worrying about it. In that car, it represented roughly 56% charge. The Mini had a much smaller battery, so nearly the same remaining distance looked very different as a percentage.

Rory says improving batteries is a good thing. His question is whether reducing range anxiety always requires adding more range.

He also describes a coffee concept that was being trialled: flat whites, with regular milk or oat milk, prepared in advance so people could tap their card, pick one up and leave. The idea was to give up customization in exchange for good coffee served quickly. Someone heading for a train might value that more than a long menu.

He contrasts this with walking into a familiar coffee shop and only then being told it won’t make the drink you want. One customer chose the trade-off before arriving. The other discovered a restriction they weren’t expecting.

His Moxy example makes the same point. He describes choosing small rooms and fewer traditional hotel services in exchange for shared spaces he actually wanted to spend time in. That works for someone who understands the offer and wants it. Someone expecting a conventional hotel could be disappointed by exactly the same room.

Before deciding that a product needs more features or better specifications, it seems worth understanding what people think they’re getting and what is actually bothering them. In Rory’s examples, the number on the screen and the expectations someone brings to a purchase are part of the experience too.

[Harper Carroll @HarperSCarroll]

It’s crazy to me that almost everyone keeps their airplane windows closed now. What changed??

Pretty sure I have the only open window on this plane

Crazy bc this view is one of the most amazing non-digital experiences we have access to right now

[Michael Fallon @_michaelfallon]

@acesaasdemos A bit of both. I’m also re-ranking and filtering by hand. Once I’ve collected enough of those judgments, I want to test whether a model trained on them does a better job evaluating and ranking the results.

[Michael Fallon @_michaelfallon]

I still find it amazing how difficult it is to get AI to produce a short summary that captures everything important, consistently. Summarizing was one of the things it seemed good at early on, but I still spend a lot of time trying to get it to do that reliably.

[Michael Fallon @_michaelfallon]

@packyM Pretty much haha

[Michael Fallon @_michaelfallon]

@packyM I've started firing some

[Michael Fallon @_michaelfallon]

@paulg And somehow there’s still time to worry that you’re not spending enough time on it.

[Michael Fallon @_michaelfallon]

@athyuttamre Not voice-specific, but I haven’t been able to get dot to see or control the browser inside another local ChatGPT chat. Would love that to work.

[Michael Fallon @_michaelfallon]

@packyM Funny that it can control Claude, but I haven’t been able to get it to see the browser inside another local ChatGPT chat.

[Michael Fallon @_michaelfallon]

@mcuban Do you ask for the source passage with every answer, or only when something looks off?

[Michael Fallon @_michaelfallon]

@nikitabier When you pulled people onto spam, how did you decide what they could safely stop working on?

[Michael Fallon @_michaelfallon]

@HarryStebbings Even if a company never switches to an open model, having a credible alternative changes its next pricing conversation with a lab.

[Michael Fallon @_michaelfallon]

@nikitabier Did the fabricator take your first set of CAD files, or was there much back-and-forth before they were ready to cut?

[Michael Fallon @_michaelfallon]

@rexsalisbury I think it’ll be like shipping containers, where most of the value accrues to the people using them rather than the people making them.

[Michael Fallon @_michaelfallon]

@klashsarma https://t.co/ivLjRnb99d

[Korva @korvaapp]

the personal growth was temporary

[Michael Fallon @_michaelfallon]

@garrytan The surprising user for these benchmarks may be human juniors. Give them the same cases and rubrics, and a firm gets a way to teach judgment that currently only shows up in a senior’s review notes.

[Michael Fallon @_michaelfallon]

@paulg It probably changes who agrees to give the talk too. Someone might be happy to work through an unfinished idea with 20 people, but not want that version on YouTube forever. You lose some interesting conversations before they even happen.