taladar @ taladar @sh.itjust.works

Posts

1

Comments

2,384

Joined

2 yr. ago

So you are saying it is too creative for the average person in marketing?

Might be time for a rewrite in something more modern anyway.

That dialog sounds like the AI version of the typical unhelpful FAQ page that answers the questions the company wants to answer instead of the ones that are actually frequently asked. In that situation I mentally tend to pronounce it as Fa-Q (fuck you) page.

Honestly, if the existing victims have to deal with a few more people masturbating to the existing video material and in exchange it leads to fewer future victims it might be worth the trade-off but it is certainly not an easy choice to make.

Which countries do you have in mind where videos of sexual child abuse are legal?

Does it feel odd to anyone else that a platform for something this universally condemned in any jurisdiction can operate for 4 years, with a catchy name clearly thought up by a marketing person, its own payment system and nearly six figure number of videos? I mean even if we assume that some of those 4 years were intentional to allow law enforcement to catch as many perpetrators as possible this feels too similar to fully legal operations in scope.

Rate limiting in itself requires resources that are not always available. For one thing you can only rate limit individuals you can identify so you need to keep data about past requests in memory and attach counters to them and even then that won't help if the requests come from IPs that are easily changed.

Yeah, the problem with enshittification is not that it is something that some companies do but that all companies are heavily incentivized to do under a lot of circumstances (enough that circumstances will come up for practically any company regularly).

Plus security is one of those if everything goes right "what are we paying you for" and if something goes wrong "what are we paying you for" parts of the business.

Also politicians should just permanently be ineligible as office holders if they bring up crypto war topics again. Like adding backdoors to cryptography "just for the good guys". People that stubbornly refusing to acknowledge the laws of nature just don't deserve the right to be voted into office.

Agreed. Even human translations are usually crap 99% of the time if they are done as pure surface level translations without having someone who understands the original in depth essentially rewrite it in the target language.

I keep wondering lately if email isn't borderline useless for a lot of people anyway when it comes to communication with actual people (specifically outside of a business context).

Aren't many of the legitimate emails we receive today from systems where we already have an account and could be replaced by some sort of system of web hooks with service specific auth tokens? Potentially the messages there could be partially machine readable too (e.g. JSON) since most fall into similar categories (invoices, delivery announcements, password resets, calendar invites and/or reminders, bank account or other payment provider transactions or transaction logs for entire months, notifications about direct messages or other events received on some web application or game,...). This might also make it easier to process these automatically (e.g. let the user tell the system to automatically save all bank statements to a specific hard drive folder instead of manually logging in to download them each month after a reminder mail).

Both in English and my native German. I probably do have an accent in English but that is difficult to judge myself. Certainly nothing that prevents other people from understanding me though.

I think the problem is that you can't create new abstractions very well in graphical languages. It works for something like fixed domains (e.g. Blender node editor or your example) but for a general purpose language you need the ability to define abstractions that never existed before.

The other problem is that you can't really apply any of the tooling to it that works with other languages, e.g. version control, formatters, linters,...

I thought about it some more since I wrote my comment and I am genuinely unsure any voice recognition system I have ever used managed to transcribe a full sentence to text successfully without making at least one mistake.

On the other hand with a keyboard I am reasonably sure I get problems such as network filesystems being unable to reach the server or broken hard drives more often than having to worry about mistyping a command I commonly use. Granted, part of that is thanks to tab completion but that is part of the issue with voice input, no easy way to correct what it got wrong.

Works fine on PC with an ad blocker and at least there you aren't relying on YouTube keep any promises.

On the other hand exclusives really just make everyone else who isn't so easy to manipulate hate your platform.

As a primarily CLI user on Linux I wouldn't even think of most commands as "words per minute" unless I am composing a complex pipeline or run a command with dozens of parameters at which point typing speed is not my bottleneck limiting the speed of input anyway and a free conversational interface would be totally fucked trying to figure out what I want it to do.

In my own real world usage I estimate a comprehension rate of about 92% with voice agents.

For me it feels more like 9.2% most of the time, and that is just the voice-to-text part, not even the interpretation of the resulting text as a command.

Maybe the tendency for LLMs to shower the user with praise for their prompts also makes them attractive to egocentric CEO type of personalities?