frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Ask HN: Positive text origin verification

3•Miguel_FierroM•5h ago
Is anyone working on a positive text origin verification standard?

Everyday it gets harder to distinguish between AI generated content and human generated content.

This will never really work on the side of the labs but for people looking to actively prove they're the ones that wrote something there's an incentive to voluntarily collect that evidence.

Is anyone working on this?

Comments

Juliate•5h ago
Why should it be to the people to prove that their own work is not AI, and not the other way around?

EDIT: I mean, the effort has not to come from people that do the work by themselves. A new technology is not beneficial if it imposes more effort than before to do the same work.

Miguel_FierroM•4h ago
I'm working under the assumption that there's still some value to real human interaction / media, which I do believe (not in a "end AI" way, exactly the opposite). There will always be an incentive to pass anything as human produced not only by the labs but by the people using these tools. As soon as Anthropic announced their watermarking strategy, there were people reverse engineering it and opensourcing tools to remove it.

In practice what has ended up happening is that in almost every surface people should assume that there's an increasing level of probability that that specific piece of media is produced by AI.

Juliate•4h ago
Of course, we should assume that, and we should assume people to honestly disclose their own use.

But what is infuriating and disloyal, even, is the suspicion that this raises on everyone's expression, the distrust toward everyone's capacity to think and to write, and to speak, because "it could be an AI that built/made/assembled all you've done".

And this is the result of releasing such tools widely: people that still do independent work are now requested to bear the effort to prove that they did it "all by themselves". "Proof of work" becomes a thing, for instance, that artists have to spend more time on (as well as the many other "visibility" taxes social media brought upon them), rather than doing the work.

In other words, tools that are supposed to help do the work, only add more administrative paper work to everyone.

GenAI taints everything it touches, not only because they all have been trained with stolen and illegal stuff, but because a lot of people are not able to distinguish anymore the difference between genuine and GenAI stuff, and the price is paid by those who can (and I say that having tested and used it).

bbor•4h ago
I think "standard" needs a lot more detail, here. To me that says "protocol", and I don't see a clear way a protocol could prove text provenance, as any protocol is ultimately only as good as what one side passes into it. Even if you were to somehow seal and stream keyboard events without invading privacy, that's just raising the bar for fraud -- a frontier model could whip up a python intermediary to simulate human keystroke speeds in 30 minutes or less.

I suppose the very hardest evidence to fake would be a video of the author typing, verified by AI (we'll soon lose the ability to distinguish on our own, if we haven't already).

Perhaps even better would be a keyboard with physical yubikey-esque technology... shit, that one might actually work. If someone has a robot dextrous enough to type, text provenance will be the least of our concerns, anyway!

I took a shot at something like what you're talking about, tho it's less "verification" and more "uncovering" information the creator is freely giving away. Undertone[1]: a local Python indexer plus a Firefox/Chrome extension that tints text by evidenced git provenance.[2] I've been too overwhelmed with the job search to finish it in any meaningful sense, but there's freely licensed scraps there you're/anyone is free to use if you'd like :)

But again, to be clear: it relies on good faith alone, and is intended for websites exclusively. Would never coherently work for non-VC'd prose (read: most prose).

[1]: What does it mean? I'm not sure. But it sounds provocative! It get's the people going!

[2]: https://gitlab.com/doering-ai/apps/undertone

Miguel_FierroM•3h ago
Yes! This is interesting. I'll have a look.

And for sure it would need either the yubikey-esque keyboard and/or the adoption of a standardized origin system that publishing surfaces can verify against.

Seems like this will at some point need to be solved, or the way we interact with media will have to fundamentally change. It is an option at the end of the day. Certainly will be fun to see what happens.

structuredPizza•4h ago
Something adjacent to what you describe is C2PA: https://spec.c2pa.org/specifications/specifications/2.4/spec...
Miguel_FierroM•3h ago
Thanks! I'll have a look.

Tell HN: I've been paying for a rural Tanzanian's education for 10 years

629•lukehandcool•7h ago•176 comments

Ask HN: What do you run on a $5 VPS that's worth keeping online 24/7?

49•mariocesar•1d ago•54 comments

Ask HN: Is the Steam Machine worth $1049?

8•bigwheels•2h ago•6 comments

Tell HN: GitHub refuses to remove cracked copies of my software after a month

560•IvanK_net•2d ago•341 comments

Arc AGI 1 400/400 No LLM running on my MacBook

2•5CEOs•1h ago•0 comments

You're saving one minute and now I can't know if I'm talking a human

9•scoofy•3h ago•3 comments

Ontological Language Model (OLM) + LLM: Deterministic AI Advancement

2•northlondoner•2h ago•1 comments

Ask HN: Any Insight Here into French High School Unrest?

3•giardini•59m ago•0 comments

Super Intelligence and/or Artificial Intelligence

4•butlersean•3h ago•2 comments

Ask HN: Positive text origin verification

3•Miguel_FierroM•5h ago•7 comments

Google Search is down

6•muazzam•1h ago•3 comments

Ask HN: What happened with quantum computers anyway?

3•pera•4h ago•2 comments

Ask HN: Do employees vastly outnumber founders on Hacker News?

4•amichail•7h ago•4 comments

Ask HN: Why hasn't AI replaced Tax filing software

6•thelastgallon•10h ago•9 comments

Ask HN: Got a Mac for running agents? what's your setup?

2•velyan•1h ago•2 comments

Ask HN: Are there AI models for generating sounds based on a text and reference?

34•onemiketwelve•2d ago•18 comments

Ask HN: Learning practical system design

2•rcdmrl•10h ago•0 comments

Ask HN: Are we trading model capacity for compute?

2•olirex99•12h ago•0 comments

Tell HN: Amazon forgot how to write and test code

11•penskymaterial•1d ago•4 comments

Ask HN: Is the Future of IDEs Bleak?

3•bigmattystyles•22h ago•8 comments

Ask HN: If AI doing decades of math now doesn't get our attention, what will?

2•silexia•2h ago•2 comments

Ask HN: What Is Your Personal Backup Methods

11•Crontab•1d ago•12 comments

Ask HN: What games do you religiously play on your phone browser?

14•saimiam•1d ago•32 comments

Tell HN: HN has lost the joy for me

35•trencedamp•1d ago•27 comments

What is the actual percentage of people for and against AI?

11•Zeruxe•21h ago•54 comments

Ask HN: How do you describe what your relationship brings you?

7•maxignol•2d ago•8 comments

Ask HN: Which frontier model can do code security reviews

6•Bender•1d ago•6 comments

Ask HN: Is the xcancel source code publicly available?

3•bigwheels•1d ago•1 comments

Ask HN: Is anyone still working on dLLMs

3•obonse•1d ago•3 comments

Ask HN: Do you still own a printer?

16•nunorbatista•2d ago•29 comments