I have been using (and contributing to) on an open source project called Zaparoo[1] that lets my kids (5-8) play retro games, watch videos, and listen to music using NFC cards. The whole thing runs on a Raspberry Pi running Batocera[2]. I program the cards using the mobile app and my kids like cutting and pasting the cover art on the cards.
All the media is local to my house- I am the librarian who curates the selection of media based on my kids interests, maturity, and my comfort. It feels like the only way forward.
This is excellent UX for feed discovery. I always found the feed subscription thing distracting- usually I am reading blogs to solve a problem or research and not collect/socialize. That is something I am in the mood for later.
I love the OPDS feature to serve ebooks to my families ereader devices. Many of the other OPDS servers are rather complex by comparison and as a bonus I can use it from a web browser for my devices that don’t speak OPDS.
I put koreader on my families various devices. I have Inkpalm 5, Kindle 11th Gen, and an older Kindle. My favorite is the Inkpalm 5 but they stopped making it. :(
Why is this the best business model we can collectively execute on? Whether it is AI, home cameras, or fridges it seems to just come back to, welp, lets slap an ad on it.
Unlike conventional businesses where a good or "binary" service (it works or not) is sold, advertising is a much more nebulous good whose efficiency can't be accurately measured. This means there are tons of inefficiencies where middlemen can skim something off the top:
* a product manager decides to include ads in some digital product. Their analytics show plenty of "engagement". The engagement is actually people accidentally clicking on the ad while hunting for the tiny "close" button, but even if the PM suspects it, they have no reason to volunteer that information. They keep getting their salary paid and even earn a promotion based on the engagement numbers.
* the developers are tasked with implementing the advertising infrastructure - they get paid while padding their resume about how they're building "scalable" systems.
* the "scalable" system runs on a cloud provider and earns them a ton of money. Cloud provider is happy.
* some marketing agency is given a budget to go and spend on ads. The person there maybe even knows that advertising in the aforementioned product is a bad idea because most of their clicks are fake... but if their client is tasking them with burning money, why would they refuse?
* a marketing person at a big company that doesn't actually need any more advertising to succeed is given a budget and spreads it across a few marketing agencies including the aforementioned one. They get paid, why should they refuse?
At every layer (and I haven't even listed them all), people get paid by skimming something off the top. It doesn't matter whether the advertising works, because nobody in the chain has any incentive to admit it while the status quo is so lucrative, so the rational thing to do for everyone is to not rock the boat.
Customers are generally low-information shoppers. They go to a hardware store and ask the salesperson for a fridge that fits their requirements. The rep will show them a few options, and then the customer gets to try them out. This is where the animal brain takes over: Samsung designs for the animal brain. It's sleek. It's futuristic. There's so many doors. It has a beverage drawer. A condiment drawer. You can customize the panels. The animal knows the Samsung fridge is better, and customers likely won't know any better if the salesperson doesn't tell them (and would they? They make a better commission on the more expensive fridge)
What the outcome actually happening is indicative of however is that consumers are very very very bad at their job (consuming the best products) and do not have enough rights.
If a customer was entitled to a working product without this kind of deficiency, and we had courts that actually applied punishments to large corporations (instead of unilaterally and without justification, significantly reducing fines to nothing) we wouldn't have this problem. It wouldn't be possible to profit off of this kind of advertising because you would be too busy signing court documents about how you suck at building stuff.
There's only so many human beings who can buy your fridge. There's only so cheap you can build your fridge. There's only so much you can charge for your fridge. But line must go up.
This is simply what it looks like when the people with money and resources decide that a stable and reliable profit is a Failed business.
I think it's mostly about squeezing consumers for more money, even after they already paid a premium, because they simply can and nobody will do anything about it.
Why do you address us as if we collectively went down to the town center and three dozen times in a row and decided on the same thing by consensus? For most of us this was shoved down our throats by sheer force of violence. And why always this oh shucks apologia about the “business model” that they are supposedly forced to adapt? No, this fridge already costs a lot of money. The ads don’t have to be recouping losses. They could just be for more profit.
Because it's a dual revenue stream. The retail customer pays you, and then the advertising customer pays you. Why make only $1 when you can make $2, $3, $4 over time?
If your next question is "why do they need to keep making more money?", the answer is capitalism.
When you get downvoted for making the obvious statement that you have to maximize profits as a capitalist entity, well, you know you’re in a venture capital forum.
It's an inexpensive revenue stream; the secondary effects and risk to customers are considered relevant insofar as they can negatively impact the company's future profitability (if then).
There's no way that this was ever /not/ going to happen under current laws (US).
Internal incentives not overall profitability drive such behavior.
An executive can point to a profit stream and suggest that’s beneficial to the company while ignoring externalities that cost the company 5x as much. Nobody inside has complete knowledge if someone was a good idea or not so the appearance of benefit often replaces the search for actual benefit.
How do you feel that the paper bill counter in your bank is closed source? It does not matter because it's trivial to verify. The counter says "here is a pack of one hundred ballots for candidate A," and if you're in doubt, you just count them again. While representatives from candidate A's and B's team are watching.
The difference is that I know the sum ahead of time and can object in the moment at the bank.
A vote recount and/or judicially called audit can take months to resolve. This can lead to a loss in confidence in the outcome and for political shenanigans (e.g. Bush v. Gore).
I feel far more confident in a system where the software is open source because it increases trust for free. As a citizen having the software be open source is only upside to me.
The comments on this have lots of folks focused purely on the software, talking about a lack of paper ballots, etc. So, let me provide some more context that is missing from the post.
For those who don't know the VotingWorks software is both Open Source and their machines create and count paper ballots. You can read about it here: https://www.voting.works/machines
Essentially they have a computer, a ballot marking device, that people can use to mark their ballot. That ballot is printed on paper. Then the paper can be validated visually. Then fed into a machine to scan and count. The paper ballot is preserved and can be later audited.
The ballot marking device has a number of advantage over pre-printed and hand marked ballots:
- American Disabilities Act (ADA) compliant using standard web technologies
- Available in applicable languages without lots of translated papers on hand
- Errors or typos in ballots can be fixed days before election instead of weeks (due to print shop lead times)
- Better UX for complex races where things like ranked choice, choose three, etc with rules which can cause people to mismark and then have their ballots rejected
- Avoids sloppy/incomplete markings that must be interpreted and judged by counters/auditors
The entire system runs offline. It is open source.
They also have separate open source software for running risk limiting audits using the paper ballots: https://www.voting.works/audits
This is an excellent overview & much needed context. I read the (very short) OP but didn't dive into other sections of the website (which is not an initiative I'd previously been aware of).
Probably a difficult task to ensure all readers of all pages on the entire website are fully aware of this context in advance - I'd imagine this kind of averse reaction will continue to be common until these kind of hybrid systems become more widespread (or the interests pushing paperless are comprehensively silenced, which seems less likely).
---
That said, now that I do have full context, I do have two criticisms:
1. Clicking through to the VotingWorks frontpage, the copy still doesn't really highlight in a very obvious manner the paper nature of the system. You really have to analyse the website to figure this detail out.
2. The homepage does contain a section entitled "Faster Election Results" - which I do think flies directly in the face of many criticisms in the HN comments here & I personally believe to be an approach that's incompatible with democratic integrity. Counts should simply not be trying to be fast as a high priority - verifying the automated count by hand is insufficient if it isn't done as a matter of course. Ideally, live, while the count is taking place. The latter is not feasible with an automated system, & the former is a lot more likely to be overlooked if speed is a priority.
We don't just need systems that can be fair, we need systems that incentivize fairness & don't contain perverse incentives - count speed is exactly such an incentive.
reply