Tech companies respond to incentives. If the technology exists to create and sell devices that are useful to people and legal, then they'll do it, and people will buy and use them. And that is fantastic - I'm great with tech companies making tools that makes my life easier. I absolutely plan on using the heck out of the next wave of AR glasses for digitizing historical documents so I can have both hands free instad of holding a phone in one. Better audio recording devices? Yes, please.
Simultaneously - I own my words, either written or spoken. This isn't really all that different from copyright, and I'm rather surprised we don't more regularly bring that reasoning to bear on this issue. If I give them to you in a private conversation, then those words are for you to hear. If you record them without my consent, then you're violating my (copy)right. If a company takes those words without my consent, then that company is violating my copyright.
I don't know exactly what should happen to people and companies violating my copyright, but I do remember when it was the other way around a few years back and the RIAA was getting people fined thousands to tens of thousands of dollars per incident. That seems about right.
> Simultaneously - I own my words, either written or spoken. This isn't really all that different from copyright, and I'm rather surprised we don't more regularly bring that reasoning to bear on this issue. If I give them to you in a private conversation, then those words are for you to hear. If you record them without my consent, then you're violating my (copy)right. If a company takes those words without my consent, then that company is violating my copyright.
The copyright situation is interesting, at least under US law, because is most cases where someone on the street records you speaking you will not copyright. That's because of a loophole in the way the copyright statute works.
People commonly believe that when the US changed from copyright coming into existence when you register your work they switched to it coming into existence as soon as your create the work, but that is not correct. 17 USC 102(a) tells us that copyright subsists in works that are
>> [...] fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.
But what does it mean to be "fixed" in a tangible medium of expression? For that we must turn to 17 USC 101:
>> A work is “fixed” in a tangible medium of expression when its embodiment in a copy or phonorecord, by or under the authority of the author, is sufficiently permanent or stable to permit it to be perceived, reproduced, or otherwise communicated for a period of more than transitory duration.
Their recording of you is not "by or under the authority of the author" so is not a fixation, so your work is not copyrighted yet.
Congress recognized this loophole and partly closed it in 1994, with 17 USC 1101, which is titled "Unauthorized fixation and trafficking in sound recordings and music videos". This closes it in the case where you are, say, a street musician and someone records your song which is not yet copyrighted because you have not yet recorded it or written down the sheet music or tab. It gives such a musician essentially the same rights and remedies they would have if they had copyright.
But that is explicitly for music. Unless you start singing everything you say in public it won't help you. Even if you do that, those devices that just immediately transcribe what they hear probably would not be covered, because no matter how creative a melody you come up with to sing "I'll have a medium Big Mac meal with a coke please" to, they will just be recording your lyrics. US requires a certain level of creativity in a work for it to be copyrightable and "I'll have a medium Big Mac meal with a coke please" is probably not going to make the cut.
You best defense, if you want to have a chance of federal copyright applying, is to wear something of your own that records you. Then your words are fixed under the authority of the author (you) as soon as you speak them, and so if they do meet the bar for being copyrightable they will be copyrighted.
There is also state common-law copyright law to consider. Federal copyright law preempts state copyright law, and with it so easy to get federal copyright you don't really tend to hear much about state copyright law.
However, that loophole that makes federal copyright law not attach also makes it so state copyright law is not preempted for works that fall into the federal loophole. State common law copyright tends to be more generous or expansive than federal copyright, so this is probably your best copyright bet.
This suggests that you may want to actually avoid recording yourself, to avoid any chance of getting federal copyright and thus having state common law copyright law preempted.
I did a little searching, and apparently whether you should record yourself or not has been debated and analyzed in other contexts like stand-up comedy performances and public lectures.
That makes sense. I don't expect copyright law to apply out of the box - but more that it could be used as a reference point in thinking about privacy issues. For example if it's illegal to record a movie at a theater, then it should similarly be illegal to record and transmit someone where they have a reasonable expectation of privacy.
It's not black and white, and there's an entire spectrum of issues to consider, but preserving someone's right to privacy seems conceptually close to preserving someone's right to control how their work is copied.
>My interpretation of OP is that they trust Wireguard’s code to be secure - random wrappers, somewhat less so.
There is not such thing as "wireguard code". Wireguard is a protocol, and "the security of wireguard" is the inability of the attackers to read the contents of the stream. This is not affected by the wrapper, because a wrapper is no more hazardous than the high seas, i.e. the Internet.
Speaking of implementation code. There are three wireguard implementations that I am aware of: Go, Linux, and OpenBSD. I am quite sure that none of those are bug-free, but adding a wrapper around wireguard packets has NO effect on the security of the implementation. The implementation is receiving packets from "something". This something can be either the net, or the wrapper, or even /dev/urandom
> because a wrapper is no more hazardous than the high seas, i.e. the Internet.
Unless that wrapper is running on the same machine as my Wireguard code - and then the wrapper becomes a potential weak point in the system.
I see where you're coming from and agree with you - in theory, a wrapper is usually thin and shouldn't impact the overall security posture of the system. But this is 2026. AI is regularly curb stomping any notion we had that we write good code and someone just discovered that my SSH server has gone the last decade allowing a login to anyone that pounds out enough 'A' into the password prompt [1].
Given that, I have a fair bit more trust that Jason's Wireguard code (the implementations you mention) correctly handles random packets off the Internet than a wrapper running in the same context.
Trivially, any layer on top very much influences the security. It usually runs on the same server, and in fact is often combined with wireguard code into a new VPN. In fact there are additional layers: firewall code, networking stack code, VPN, rest of OS. Just recently there was a critical vulnerability in Linux kernel TCP connection handling code. It wouldn't matter if OpenVPN listens to TCP connections on the same server: attacker may become root before meeting the VPN.
You can write a bare wapper in few hundreds lines of python, but it won't be robust to resume connections etc. The next step up is something like ProxyGuard referenced in this post, but look at repo to see the size of the code. It's this software that listens to open port, and if it has vulnerabilities, there will be remote code execution on the wireguard server. Then we have more featureful software such as wstunnel that are much bigger.
Most if not all software changes with platform. That's trivial.
I have dealt with this issue for a while.
Update: It seems you acknowledge the issue when you say in another comment "as long as the remote attacker does have sudo privilege". First, remote attacker may get root. Second, even if someone runs as a non root user on the same wireguard server, that's equivalent to that person bypassing the wireguard plus logging in as that user, which is worse. Authenticating to the server is an additional step that they would have to do even if they bypass the VPN.
Has anyone done any work on what it is that makes LLM Voice hard? Something I've noticed, and have seen other people comment in passing is that reading LLM output is "exhausting". It's often attributed to trying to detect the machine, or filter out BS or word density or some other explanation, but it feels like something more than just that. Like - there is something about LLM output that just doesn't "fit" our minds correctly, even though words are in the right places.
Surely someone with a neuro/psych background has picked this up and developed an evidence-backed explanation.
Compared with human writing, all nine frontier and near-frontier models use more inanimate subjects with animate verbs and more low-frequency n-grams, and most also use longer dependencies, more stacked nouns, and more reduced object relatives in conversation. [...] We propose and test a hypothesis as to why: while human language is optimized for successful communication, AI-generated text is increasingly optimized for other goals like math, code, and reasoning, in which there is a pressure for information density rather than successful communication.
I read a lot of old material from 1-1900AD where I can pick upon the author's 'voice'. There are subtle clues, like if they had a regional accent, where they use a specific word or type of cadence.
AI does the same with responses where the cadence is some type of formula that becomes identifiable.
Each of the big models has their own but an example of the cadence is
A) Agreement or Acknowledgement with the User
B) Specific detail gets a paragraph of detail related to the previous response
C) Overview of the conversation's current state
D) Prompting or Direction for this conversation to go
E) Landing on a beat. Wrap up the vibe. Recap with a zinger.
I cannot unsee it and a few of the models are annoying. I have attempted to build out the Agents and context to avoid the 'last beat one liner' without success because after 2-3 more turns, I get "That challenge was accepted and you stepped up. You're the winner this round".
The 'zinger' / prompt might be injected by the harness, and not under control of the model. I am pretty sure this is the case for gemini.google.com - it's glitched a few times for me in such a way that it's revealed special markup right before or surrounding that leading follow-up question.
Because this isn't a library, but an ambitious commercial headed [1] tool. If your needs and use cases match the feature set they're aiming for, and their market research and projections are able to undercut Adobe by an order of magnitude, then they might be on to something.
> Strategic Vision: Evolve PaperPilot from a best-in-class PDF workstation into the Universal Document Operating System for humans and AI agents. [2]
Wow, a 155 page book for a relatively new project is interesting - what was your workflow for that? It looks mostly generated - is it used in its entirety as context to a top-level agent when working on new features, or are you directing it to just take relevant subsections? Why the PDF output (it does look nice)?
I do not use book for a context, book is the outcome of the development. I am still old school I feel any product should have good documentation and should be available offline, also personally I like physical things so many of them will print and read. I follow a rigorous process I used to follow since 1998, without AI, which is write something like PEP or RFC before starting work on things. So if you look at the whole project it has design records and each contains details. I read through the docs, even though I use AI from time to time. I used to write a lot of restructuredtext (RST) docs. So I wanted to move the whole pipeline to rust, I used to use latex for it earlier but then wanted to move to typst. So for it wrote another system which helps me with website, html, pdf. You can take a look if you like the project is https://github.com/insanai/guidedog, sibuna also I ported from typst pipeline to guidedog which uses rst markup. If you want to look at development part I especially use dsc (short for discussions) so if you type `guidedog dsc --help` to see the details.
> There is a palpable public hunger for historical perspective in the 2020s. How could it be otherwise? Most citizens grasp very well that they are witnessing changes of momentous significance right now and that historical change of a profound kind is taking place in our time. Yet this moment finds the academic profession of history ill-prepared to benefit. Many historians experience their discipline as being in terrible crisis. Thousands of promising graduates are unable to find sustainable positions, while those who do often have to work under institutional constraints and pressure from administrators or colleagues, making them likely to postpone or even abandon the kind of rich, bold, innovative synthetic work that would help refresh the field and reestablish its intellectual centrality to understanding the world in the eyes of the public.
So, academia has become preoccupied measuring metrics and (temporarily) lost perspective: history isn't about the past - it's about the future.
I would argue that in this moment, history is too important to be left to just the specialists. We're not "witnessing change" - we are actively making change, and there is a lot more to be done before we're anywhere near equilibrium again. We can't go back and repeat the past (nor would we want to), but we can make better decisions about our future if we're aware of it. The generations not long before us went through a world in significant change - and now it's our turn. We should learn from their experiences.
Simultaneously - I own my words, either written or spoken. This isn't really all that different from copyright, and I'm rather surprised we don't more regularly bring that reasoning to bear on this issue. If I give them to you in a private conversation, then those words are for you to hear. If you record them without my consent, then you're violating my (copy)right. If a company takes those words without my consent, then that company is violating my copyright.
I don't know exactly what should happen to people and companies violating my copyright, but I do remember when it was the other way around a few years back and the RIAA was getting people fined thousands to tens of thousands of dollars per incident. That seems about right.
reply