iiJDSii

What is something you or someone at your company has to do repetitively on a desktop?

For example, it may involve some minor variations, could be kind of 'messy' and involves clicking through various programs, checking outputs, deciding from there, etc.

Example: during an internship I remember running regression tests, which meant running some scripts, re-running them if they hanged for too long (don't ask), checking output log files for a few specific lines, and then emailing the results to my supervisor. Having to babysit the software run and do the manual stuff at the end was a pain and seemed like a waste of time but was unfortunately necessary.

Would be interested to hear other specific examples, as me and a few buddies are working on some desktop agent programs to try and tackle some tasks like this.

Are there any datasets, with any labels or no labels, for people using their desktop computers to do mundane tasks like open applications, browse the web, interact with spreadsheets, etc?

I was thinking this type of desktop "screen record" data could be useful for training AI Agents in the future.

It seems the hardest part of getting AI to "just do work" on my laptop is being able to navigate different applications, copy and paste stuff, trouble shoot weird one-off interactions like software updates moving some buttons/features to different places, etc. Rather than trying to programmatically figure everything out from e.g. an excel API to do spreadsheet work.

Thought this could be an interesting way to share potential ideas, but also discover potential products. Last time I mentioned a software I wish existed to a friend; they told me about a specific product that pretty much did what I needed.

So...

What software product do you wish existed that you'd use?

Interested in what may now be a somewhat contrarian idea. It's startup school 101 to avoid "SISP": Solutions in Search of a Problem. I think YC's startup school actually coined this acronym.

But I'm wondering, have great businesses and products started from this? There's plenty of downsides that have been discussed, but I'd like a different perspective too.

One could argue that anything that began organically as some techie's for-fun side project might fall under this category.

I'd even argue that something like ChatGPT, created by a (formerly) non-profit org, might qualify as a powerful example. I mean, a computer program that just talks back to you and answers generic questions? Sounds like a cool tech project, or 'solution' on the surface, without a specific problem in mind. Of course, now we see it has a transformative technology that at least has a monthly subscription business model that many people are willing to pay.

How about the VR headsets? I doubt the inventor(s) sat down with the explicit goal of solving a problem after extensive market research and customer interviews.

My initial take: on the surface I think this is a "high risk, high reward" approach to technology entrepreneurship. If you take the opposite approach, problem-focused, iteration-driven, lean startup, etc, you're far more likely to end up with something like payroll or accounting software. Not that there's anything wrong with stuff like that, it's very important in its own right to solve more mundane problems well.

Curious to hear your thoughts.

Curious as to people's experiences with Crossover. Has anyone hired great candidates from there, or been hired themselves at a good position?

Or maybe someone's had a bad experience worth hearing about?

I'm curious as to what unique value add is being done on top on the core chatgpt (or claude or whatever) website and chat interface.

Now they have very long context windows, ability to upload and parse through files, custom system prompts, can work with images, remember user history across chats...

So what's the unique value adds in other products that are using chatbots?

My 'hot take' currently is that the core product of chatgpt and similar bots is so good that there's very little value to 'wrapper' apps/products. I'm a heavy tech user and all of my use cases and problems I can solve with the core chatgpt right now.

The only real limitations and problems I can think of are:

- privacy, though there seem to be modes for not using/storing your data, depending on your level of trust of the companies, - reliance on internet, I suppose one day I'd need a local chatbot with gpt4 capabilities somewhere remote, - maybe file uploads (and RAG capabilities) are limited on the core gpt4/claude sites, though I'm not sure this will remain the case for any length of time.

So yeah I'm not sure what unique value 'wrapper' products are providing, but I'd love to see examples and be proven wrong and hopefully learn something new.

Can you think of any majorly successful products or companies that launched and were pretty much ignored?

What happened / what did the founders do to eventually get traction?

Let's say you have an AI Agent that can control your computer, or do some computer actions in a limited environment. Or maybe control some other hardware device!

What would you like an AI Agent to do for you? Why would this be valuable for you?

As an unknown/nobody with a startup and a pitch, have you personally gotten anywhere cold messaging investors? Emailing them, messaging on LinkedIn, filling out website applications, etc.

What were the results, and how many did you have to contact before you got anywhere?

How do you know when to stop working on something?

Context being something like an early stage startup, or side project, or related initiative.

TL;DR - what I can do in 10 minutes on a desktop python app (windows or Linux, both worked fine) seems near impossible as an Android app.

I have a simple application to prototype: take a wired USB webcam, display it on the screen on a computer device (ideally a small device/screen), and draw a few GUI elements on top of it.

Using Python scripting and OpenCV, I had a working cross-compatible script in 10 minutes, for both Linux and Windows.

Then I realized I'd love this to work on an Android phone. I have devices with USB OTG, and a USB-C hub for the webcam. I confirmed the hardware setup working using someone else's closed source app.

However the development process has been awful. Android Studio has so much going on for a 'Hello World', and trying to integrate various USB webcam libraries has been impossible, even with AI assistants and Google guiding me. Things to do with Gradle versions or Kotlin versions being wrong between the libraries and my Android studio; my project not being able to include external repos via dependencies or toml files, etc.

In frustration I then tried a few python-to-android solutions, which promise to take a python script and make an APK. I tried: Kivy + python for android, and then Beeswax or Briefcase (may have butchered names slightly). Neither would build without extremely esoteric errors that neither me nor GPT had any chance to fix.

Well, looks like modern mobile phones are not a great hacker's playground, huh?

I guess I will go for a raspberry pi equivalent. In fact I already have tested my script on a RPi and it's just fine. But what a waste, needing to use a new computer module, screen display, and battery, when the smartphone has all 3 components nicely set up already in one sleek package.

Anyways that's my rant, wanted to get others' takes on Android (or smartphone in general) dev these days, or even some project advice in case anyone has done something similar connecting a wired webcam to a smartphone.