Just reviewing the code

I was warming up to this. It certainly is better than vibe coding, but I’ve started to see libraries [with this ai policy] push out bigger releases faster with bad ergonomics and unnecessary, undocumented breaking changes, because the ai quietly migrated the internal tests/examples.

Clearly, they’re reviewing large amounts with an eye on new features and they’re not thinking through ergonomics: spending time with it, writing it themselves, and seeing where the sticking points are.

Maybe if you’re using ai for the migration, then-- problem avoided, for now. As time goes on I could see it compounding and becoming a tangled mess that’s not only inaccessible to humans, but requires more and more compute to navigate than thoughtful, handcrafted code.

Dang. Their online filter bubble must be strong. Likely the kind of person who also doesn’t care about surveillance capitalism because “it’ll recommend good things to buy.”

You’d be surprised. Data centers seem like a big issue. My boomer parents hate data centers, as do a lot of rural voters who have them nearby or proposals nearby. But to your point, I also see a lot of people who are aware of problems around ai/data centers/billionaires still using ai.

Used this all the time until I recently stumbled across the cli program units

As for the power/environmental impact, for what they do LLMs are actually very low impact per-request.

Worth noting that a request is often dozens of requests now that there’s “reasoning,” even for search. As I understand it, a model will take a question , figure out the context (one request), reform the question so it yields better results (another request), if it’s doing a web search there’s requests for each result, another to compare, another to check if it answers the original request, if not it loops and does it all over again. So one request is easily, and often, dozens of requests. This is one way Ai companies can say to investors, “see, look at how much usage has increased.”

Also, we need to factor in the power to scrape and train each of those models, build the datacenters which is near impossible as these companies are not transparent about it and actively try to obstruct investigations into it. Then there’s the redundancy of all these different companies competing and doing roughly the same thing at the same time, as fast as they can, so it’s orders of magnitude inefficient energy consuming before it gets its first user prompt.

Comparing it to other assaults on the environment is not only hard to do, but a case of “the worse negates the bad” fallacy.

Reminds me of a quote I just read:

The system is responsible in proportion to the degree that the people who make the decisions bear the consequences.

- Charlie Munger summarizing Charlie Frankel summarizing the 2008 financial crisis.

+1 for SP. Started using it recently. It’s okay. Beat out other ones I was considering. Not blown away, but functional. Can sync it using webdav.

Could be an issue if sleep masks became useless due to glowing eyelids.

Not impossible. Just very difficult. Source: I lived there for several years without a car, moved close to where I worked (twice), biked and walked most of the time, some public transit when the routes lined up, had very little social life. If there was a special event across town, could use some money saved from not having a car to spring for a taxi.

I disagree. Anyone working in a skilled profession can attest to the number of people who come to them saying that they have this “great idea” for you to work on, for free. It’s never a great idea. Skilled professional have their own great ideas, and that’s often the motivator for acquiring new skills to make those ideas possible.

Lowering the bar between idea and implementation, in a capitalist, surveillance/attention-based, target-marketed world further incentivizes low-effort, high-engagement practices and likely shrinks the space for high-effort markets.

Meanwhile the tools “lowering the bar” are rent-only or need hardware that is quickly becoming unaffordable. Silicon Valley’s only play is enshittification. If it actually is successful it’s only a matter of time before they switch from growth mode to profit mode and it’s google search all over again but this time you’re paying a monthly fee. If you want to have an idea of how that works out ask an adobe user-- and those were products that were already profitable. There is no way people are willing to pay the actual cost of ai, it will need to be extracted. The rest is externalities.

Generally agree, especially non-spying, but the context here is generative ai. So if I misunderstood the message, I’d be curious to know how.

Seems to imply people don’t use Ai to shortcut their thinking and people don’t think when they work.

I think Qobuz and Bandcamp have the ability to stream what you’ve purchased, as well as downloading to own. So in the interim you can still stream, and the option to self-host or just have it on your local devices is a bonus that could be explored later.

Yeah, people have their own threat model and trade-offs they’re willing to make for security/convenience and it’s probably a lot lower for most people than if they knew more about the landscape and gave it careful thought. It sounds like other mobile devices might reencrypt after reboot. IDK about the pw v. bio, though, but if it’s like Grap.OS just convincing them to reboot their device before going through TSA checkpoints or other areas where their device might get confiscated may improve their security with minimal effort.

From what I’ve gathered, I think you’re right to assume it’s legally safer to refuse to unlock rather than wipe. Also, worth noting that you can be compelled to unlock with biometrics, but not a password. On grapheneOS this means simply shutting off/restarting your device as it requires a pw after a fresh boot, even if biometrics is enable.

RE “If you’re not paying, you’re the product,” also doesn’t always apply when it comes to not paying EG: Linux and other FOSS products.

Not that I know of, unless self-hosting. I’ve gone back to ripping CDs and sometimes buying DRM-free HQ audio. Probably getting my purchases profiled, but at least I can download it and no one else has to know when, what, & how much I’m listening to.

For easier access I have a Jellyfin server on my home network. I plan to switch to navidrome or similar for a better, dedicated music experience as well as setting up secure access over the internet for me and my friends and family.

For those not in NJ trying to avoid this, pay in cash and punch in a random phone number at the register.

You may need to try a few numbers at self-checkout to find a working one, but it’s surprisingly easy. Rotate the numbers to dirty up the data of these parasites.

Can’t help but imagine this would result in having to “read” and sign a 70 page ToS and 30 page Privacy Policy before we can buy groceries.

Beyond the obvious problem of paying but not owning, it’s impossible to watch/listen without that data being stored on a massive corporation’s server and likely sold to hundreds of advertisers. Last I saw Spotify had over 300 venders they share data with. I’ve seen sites with over 1000.

So you’re not only paying for something you don’t own, but you’re also paying to have a larger profile in the surveillance capitalism machine.