Startup Ideas

Physical products? Do they still make those!! ;)

Agreed, I feel your pain. I would think this is only initially a solution for specific use cases, SaaS being one of them...and luckily the one I do. And in a highly regulated, safety critical sector, where testing every aspect of the software to not only the level I need, but to the level the regulators demand, is what it takes.

That said, a contract doesn't necessarily imply perfect precision. You can still deliver to vague specs and ambiguous goals and still get paid, so long as you encode something that both parties agree to. I have signed many paper and e-sign contracts that resulted in money transfer the traditional way for software that never did exactly what I expected, but eventually everyone was happy. I paid or got paid.

As to your specific use case, sure, the customer has to be electronic enabled in some way. Lots of hand waving here in the spirit of ideation... but when I was in India last, for example, even in the villages outside the urban centers, most small shop owners and vendors had mobile phones. I presume similar in all corners of the world these days, but certainly wherever HN is read. So if I am selling a physical widget, why can't my smart eth app send you a smart contract quote/invoice, you approve the details and shipping, and then fund the eth contract. Payment proceeds based on 1) physical placement of the widget in the hands of my bicycle courier who also has a smart phone and snaps a photo of the widget that gets uploaded for to the "oracle" app, fulfilling my end of the bargain, and 2) the physical delivery of the widget to your chaat stand where you send a photo of the delivered widget to the app before the courier signs off.

Would that not work, even for a physical good?

Regarding the firefighting idea: why can't evacuate everybody, then have drones hold up a temporary "tent" around a building, then fill the whole space with a non-reactive gas? Such a system could be set up and work in minutes, and also prevent the usual huge water damages.

Totally agree that AP is designed as interest free financing in most companies, large and small. Though I don't know if a squeaky wheel solves that, even with AI or social engineering, etc., precisely because delay is explicitly incentivized.

I concede that you are in the "squeaky wheel" business, so you probably have data showing me it does solve that!

But I would argue in some/many cases, not your cases naturally, I think the squeaky wheel is a counterproductive social norm that actually signals I am OK not being paid. It tells the deadbeat customer, that 1) you haven't gone out of business yet so they can still delay and 2) when things get serious you will not be so subtle, and will start threatening real punitive action. Until then continue delaying.

That's why most companies insulate the AR from the actual people involved in negotiations and selling. You can always blame those "people in accounting" and over drinks promise to "look into it" when all the while you are telling that poor chap in accounting to sit on the invoice. Bad form I know, but happens all the time.

I agree that if you have a special "insider" relationship curated with the 1 person in accounting that cuts the checks, you can sometimes circumvent the accidental delays.

No one ever got a larger bonus for showing FASTER cash outflows in a given quarter. So as long as that is true, line managers with PO authority will not be squeaked into submission.

I think the trick to getting companies comfortable with "smart"-er autopay is to show them a compelling reason why it works for them on the other side of the equation. But since you are actually out there working hard in this space I'll definitely defer to your experience. Maybe it is impossible, at least in the short term. Being married to a PhD in psych, I know that changing human behavior is SUPER DUPER HARD...of course she'd say the same about my domestic behavior ;)

Oh dear, the secret of Y-Combinator's success is finally revealed - blind luck! ;)

Real ideas start with real problems.

Just off the top of my head...

1) Have you ever tried to debug an HPC program running on 700,000 cores? You are probably using a patched version of GDB and it's darn painful!

2) Why are we still using a 1970s designed programming language in embedded systems that almost guarantees baked in bugs around memory management and multi-threading?

3) What do we do about data strangulation by the big players that is going on in the web right now?

4) Why are we using presentational markup for the web rather than semantic?

5) How to bust up a travel industry that is basically a brutal cartel of the big players?

6) And it would be really, really nice if I could get my repeat prescription without having to phone both the surgery and the pharmacist, then drive to town, pay for parking, find they can't issue my meds because the pharmacist went off to do his xmas shopping, and wait for an hour until he gets back, and then have to repeat these shenanigans the following month...

7) Why do so many applications, especially those in Python, fail to take advantage of multiple cores, and what can be done about it? We live in a multi-core world now folks...even embedded ARM chips are multi-core. Related: how to solve the conflict between CPU-bound and I/O bound threads?

8) Why is there no really clean, simple way to write decent, scalable web applications? Fotango tried this many years ago with Zimki and they were on to something until Canon closed them down. There choice of JavaScript was not ideal but the best choice at the time.

9) Why, when I am planning my next trip to Philippines can't I put in my start point, and an end point and have the system generate a sample itinerary for me? If I say "yes" my accommodation should be booked automagically for me according to my budget preferences. As well as a web app there could be a mobile app that points me to my next port of call on my journey.

10) Why oh why do we insist on using a general purpose, multi-tasking, multi-user desktop operating system in applications it was never designed for - and where a single user, single process, text-mode, lightweight, multi-thread OS would be far superior?

11) Why is it when I book a flight online I feel like I'm being sneakily ripped off? I would love to see more transparent pricing of flights, and legitimate regulated price comparison.

12) Why is it that when you see a speaker give a presentation at a conference they almost always have to struggle with the presentation system for five minutes before starting?

13) Why is the US system of skilled migrant visas in such a sorry-arsed state - and why aren't tech companies working together to get this sorted? Even a basic site listing which companies would actually be prepared to fill out the relevant form to sponsor you for would be a good start. Many big companies are very vague on this point (I won't name and shame).

14) Why is the teaching of practical philosophy so shockingly bad? Book of Life is one of the best sites out there on this, but could do with some stiffer competition.

15) The baby bidet idea I liked - what comes out in the first few weeks after birth can only be described as "alien toxic goo". To go with it, a mobile app that, by emitting a cunning and mysterious range of coloured lights, weird music, and bird-like tweets sends them off to sleep - I'd have bought it - AI be damned!

... etc etc etc

Do any of these suggest possible solutions that solve problems? I'm not saying the above are in any way potential business success stories, but hopefully they point towards real-world problems.

I have a notebook filled with hundreds of such problems...

(p.s. finally, why is it easier for me, and more satisfying, to keep a paper notebook of this stuff rather than electronically?)

Prototype it? That sounds crazy enough to test out.