This would not only allow consumers to continue using the product they supposedly own, but will also cause investors and managers to pause before developing and marketing an exclusively cloud-based device. CPU and memory are good enough that most regular consumer devices shouldn't need an internet connection
Your idea may work for technical systems used by technical users, but in this case (and I'd bet my bottom dollar most cases are similar) that the products are intended for non-technical end users.
How is some random family supposed to take the public repository of source code and set it up so their child's emotional support robot works? How is your middle-aged neighbour supposed to set up their Nest doorbell with someone else's server and establish trust? How is your aunt supposed to set their Alexa speaker to a self-hosted server? These are people who call a desktop computer a 'CPU'. Put yourself in their shoes, their IT skills are basically 0. Where would they host the compiled code? How would they even know where to start?
The intention is certainly good and it would be an improvement, but it would not allow consumers - most of whom are buying an appliance - to continue using their products. Would it really also push investors and managers to pause? It's a non-monetary impact and I can't imagine they'd allocate any effort building systems that package it up so that it's simple to stand up the cloud services. It has no benefit to them.
Most consumers aren't hackers/developers. Most are non-technical. Implicit in the purchase of a cloud-connected appliance is an expectation that some geeks somewhere make everything Just Work for them. They're not looking to have to tweak (digital) knobs and configure parameters - that's the company's job. That's why so many of these sorts of devices are designed with very simple user interfaces.
Your suggestion is good within the narrow window of technical systems for technical consumers, but this is a far tougher nut to crack than that when considering the broader audience.
It’s a nice-to-have if you could resolve the part about needing to trust some random person picking it up and modifying it, or hosting their own server for you to use. Unlike when you purchase an item from a known legal entity, there’s way less assurance about your data, security, etc.
The lay-consumer isn’t going to spin up a DO droplet to host their own instance. They’d be hard pressed to sign up to ChatGPT and drop in an API key.
These are all very hard problems and I don’t pretend to have any answers.
I honnestly wouldn't even go that far. Systems that were designed for self hosting can be difficult to deploy on your own. A system that was managed by multiple teams might take a man year to unravel.
My classic car is no longer supported by the defunct manufacturer, but there is a whole ecosystem of parts and services companies who come together to support it for me. If we had better access to the internals this could be the case for expensive electronics too.
Can’t release those publicly, so you’d need someone to hold the domains, the certs, they private keys, and operate the servers. That costs money, so they’d need to collect payments and manager customers. Starting to look an awful lot like running the company again.
You probably also need the customer database including user accounts and user data. Obviously can’t release that, so you’d need some process to release everyone’s data individually on some ongoing basis.
And you couldn’t release any code that was licensed rather than owned, so users couldn’t build most software anyway without acquiring their own licenses for that 3rd-party code.
There are so many problems with this idea that would prevent it from being viable.
Eventually network connected shit will piss off enough people to do a combination of all of above without the need to opensoure and deal with it yourself.
Forcing companies to forfeit their IP to public domain is an idea that appeals to tech people who imagine they’re going to reboot the entire service themselves, but it wouldn’t actually solve these problems in real life business cases.
It’s also effectively an appropriation of the company’s IP. Any country that introduced legislation that mandated companies run products for 7 years or completely forfeit all of the IP would immediately see development and production of that product move to other countries without such laws.
These ideas get floated on Hacker News but they’re completely unrealistic.
> you can go after the owners/executives if they're not compliant.
Holding startup founders personally liable if their product or business fails before 7 years and their business can’t be open sourced would create a very bad business environment, to say the least.
This also ignores the real problem that parts of source code are frequently licensed, not owned. Companies can’t just open source everything because they’re often using some licensed software or libraries.
This entire plan would make it effectively illegal for companies to develop or release a product which contains code they licensed from another party.
The number of issues with these kinds of ideas are numerous. It’s the domain of Internet fantasy, not actual policy.
Regulation can require that sublicenses extend to private individuals for repair of devices or services that are EOL. It could even limit the versions covered and commercial use.
Regardless, perfect doesn't need to be the enemy of the good. As it is consumers are increasingly powerless and landfills are overflowing.
Any country that did something like this would see every hardware company immediately relocate to a different country that didn’t have such weird laws.
Regulation like this is how you completely crush a country’s tech industry.
> landfills are overflowing
Tech waste is a tiny fraction of all garbage. People throw out more volumes of household trash in a week than they might discard tech devices in many years.
> Holding startup founders personally liable if their product or business fails before 7 years and their business can’t be open sourced would create a very bad business environment, to say the least.
Maybe. That's a conjecture; a hypothesis at best. At the same time it might introduce pressure to have the licensing craziness tamed and streamlined. It's way too complex and hinders innovation, as you half-alluded to.
I think you have a fundamental misunderstanding of how bankruptcy works if you don’t think the IP is important in bankruptcy.
> Maybe. That's a conjecture; a hypothesis at best.
It’s not conjecture, it’s simple fact.
If you introduce laws making founders personally liable if their product fails within 7 years, you create a hostile environment for startups. It’s very simple.
Citing legalese is not relevant here, we all know how the situation came to its current state, the real interesting question is how can things improve.
If you introduce a law like this, you effectively remove that IP from the company’s assets. Tech companies are valued largely by their IP.
No sane investor or startup will headquarter themselves in a country that has a law forcing their IP into public domain.
Investors won’t invest in and banks won’t loan companies to develop tech IP if there’s a law on the books that says they can’t sell that IP.
> The legal aspect is not interesting to me.
I can see that much. :)
The legal aspect is all important. Waving it away is ignoring the entire issue.
> the real interesting question is how can things improve.
The proposals so far would do nothing other than discourage development of new products in that country. Hardly an improvement.
https://www.epi.org/publication/employers-steal-billions-fro...
Many products are developed with software licensed from vendors, which cannot be open sourced by the startup. They don’t even own the IP, just a license to use it.
There’s also more to a product than just some source code. You’d need the cloud domains, the private keys and certificates, and other bits. Releasing many of these to the public domain wouldn’t solve anything because you’d need someone to operate the cloud servers at the known address and trust with the private keys and certificates
All of this quickly begins to look a lot like requiring someone to operate the business again, which is obviously a silly thing to require in bankruptcy proceedings.
Forcing companies to release their IP if they go bankrupt is equally silly, because it renders the value of the company’s IP as $0 before anyone can be paid out.
The only thing this would incentivize is for companies to either relocate to countries with sane laws, or to shut off the servers and avoid bankruptcies by keeping the company technically alive but doing nothing other than staying as a registered business. It would act as a holding company to keep the IP in case someone wanted to buy it.
> CPU and memory are good enough that most regular consumer devices shouldn't need an internet connection
https://moxierobot.com/products/ai-robot "emotion-responsive HD camera and GPT-powered AI"
Yet, this company failed to secure additional funding at the last minute in an atmosphere where Gen AI is the buzz word everyone's after?
Is this a one off case specific to this company or is it some sudden realization among businesses about the current viability of AI products?
Been more than a year, still waiting.
Sure, it might take 50. But the AI craziness is made no less crazy in the meantime. Way too much time and money thrown for fairly marginal benefits.
Marginal benefits? Even just Github Copilot itself is absolutely worth it, itself probably a billion dollar company, and changed the world.
Image generation is another. I personally know a business owner - it's an online grocery store - who used AI generated pictures to reduce costs by 10s of thousands of Euro, it's what made the business possible. (The pictures are cartoonish, nobody thinks it's a real photo)
I have only seen it introduce friction and keep seniors busy code-reviewing the subtle mistakes that the so-called "AI" does.
That it does not happen to you, means it never happens to anyone else?
One anecdotal evidence does not nullify the other, and vice versa. You have your experience, I have mine. I even qualified my statements -- one of the languages I use is not very popular and thus statistical models like the LLMs obviously don't do well with it -- but you are happy to ignore that and keep arguing that your experience is the prevailing phenomena, which I'll always disagree with.
Peace.
Something like that could be done here. If you make products that don’t last a reasonable amount of time or you pack up in the middle of commitments then there’s some way for your relevant corporate assets to be transferred to another entity that will take up support.
I'm sorry, but the legislative branch isn't of or for you. It is of and for the companies that pay for the congressional campaigns.