I really need to get my POC out of beta and out in the world. This UI is polished, but there's several issues with it in terms of usability and extensibility and following the source of knowledge when it gets a large amount of sources
I think you'll have a few more iterations until you figure out that going with this visual graph style isn't optimal.
We are iterating fast and solving issues, indeed, and are aware of some problems.
It would be helpful to understand what you mean more precisely with the last line, particularly if you've experimented with this graph style and found it seriously inhibiting. Our assumption in fact isn't that the visual graph style is optimal for a LOT of depth. It is useful for some 'unknown' period of exploration. We have alternatively a tree representation of the same information, which we find becomes quickly better to use, once you have a lay of the topic and are more familiar with top level ideas. Then for specifi
“””
Our analysis reports that the regulations that are currently being discussed and likely to have an impact on the following services of yours: {{SERVICES}}.
These regulations touch on the following points: {{POINTS}}.
Here are how each of your listed services would likely be impacted by the different components of the regulations:
I heard about curl-impersonate yesterday when I was hitting a CF page. Did something else to completely bypass it, which has been successful, but should try this.
Not sure what you mean specifically, but generally the organisations doing screen-scraping¹ would prefer to use compliant APIs as they don't require anything like as much maintenance (bank adds a button to the login flow? Kaboom! Integration is broken...) or resources (e.g. running headless browsers).
Some markets are pretty much exclusively compliant - I don't think there are any Nordic banks that don't have fully PSD2 compliant APIs for example whereas, if I remember rightly, the Spanish banks were all over the place. I'm fairly out of date though, so things may have improved or exceptions for scraping expired.
¹ Note that I'm talking exclusively about banking integrations here, not AI nonsense.
Yes, but they aren't very detailed to a tactical, day-to-day level. He talks more about the strategy, how to choose the problem you're working on and etc...
They are very good, just not that down to earth for regular people.
How would this function in two-party consent states like California? My understanding is limited, but from what I've read, this might still violate consent laws unless explicitly disclosed—even in public spaces.
I recently explored building a real-time STT system for sales calls to support cold-calling efforts. However, the consensus from my research was that, even if audio is streamed live without storage, consent laws could still present significant hurdles.
I think it really depends on the legal definition of recording or what it's used for.
Common sense says that a recording that only exists for a few seconds, and is utilized only by the person a speaker is intending to speak to, and is never permanently stored, should be fine. And we can assume Apple has made sure this is legal in its home state of California.
But EU law might not have sufficient legal clarity on this if it was written in a particularly open-ended way.
If you put this in the data processing agreement and make those available ("GDPR" requirement in most countries) it's pretty much fine. You can probably use them for training and improving purposes if destroyed after a short duration as well. Speech is personal information but not medical so no special things apply. The laws in most countries are quite clear about non medical personal information and translation is a good grounds for usage of personal information as far as I can see.
You have to specify usage however. Which is what most companies bark about because they want to store things for later use and unknown purpose. Which is frowned upon. You also need to protect personal data adequately which is deliberately vague. Storing it unencrypted for instance is not considered adequate. This also applies in transit.
Is this much different than a hearing aid, even in technical detail? A hearing aid will have a small internal buffer containing processed audio. I’m not sure the law will care that the processing is more substantial as long as it’s on-device and ephemeral.
> How would this function in two-party consent states like California?
It won't. Regulations permit sound recording, as long as it's not stored. Speech-to-text and hearing aids for disabled people are an example of permitted uses.
The Constitution has nothing to do with this. It constrains the Government, not private actors. And there’s no Constitutional right to a translation service.
You’re probably thinking about warrantless recordings of conversations and the reasonable expectation of privacy requirement. This doesn’t apply here.
The Constitution can definitely restrict the ability of the government to pass laws regulating behavior, and two-party consent statutes are definitely laws that regulate behavior. Whether the object level question is true is a different matter, but I would assume so given that you can point a camera that records audio at people in public at all.
Also, the US Constitution does constrain private actors, all the time. It bans slavery for a very simple example.
The Constitution does not forbid laws that require two-party consent to record a conversation. That’s what we’re talking about here.
The constitutional view of the 13th Amendment is that it withdrew from government the power to promulgate or enforce laws that allowed slavery to exist. If a slave escaped after the 13th Amendment passed, the government then lacked the power to assist the former slaveholder in capturing and returning him. Similarly, without the power to enforce property rights, there became effectively no property interest in a slave.
I’m finding a lot of literature from NGOs specializing in US jurisprudence saying the first amendment has been interpreted to protect public, obvious recording[1].
If I tried to draw up an indentured servitude contract, what federal or state laws would explicitly forbid enforcing it myself? If state laws, do all states have equivalent laws?
The First Amendment protects you against prosecution for filming the actions of law enforcement in a public venue. AFAIK it hasn’t been interpreted to void state laws that might forbid someone from recording a conversation between private individuals who don’t consent. The link you provided says as much.
How do you self-enforce an indentured servitude contract? Or any contract, for that matter? Only a court can compel performance or restitution for a breach of contract.
At the point where you enable this feature (you wouldn't walk around with it enabled at all times because why?) the phone shows a screen asking you to get consent and the other person touches yes/no and that's it? Or would a signed form with a government seal be required?
IANAL, but from my understanding the user needs to get consent, not Apple. There would be no consent screen, apple would at most give a small dialog warning to the user that this usage is illegal (for the user). unless every participant has given consent
Here’s a fun one I given many years ago: I had a friend/client who was professor. We’d talk about ADHD, issues, and other things. One day, I came to him saying, “a lot of times, I’ll read a paragraph 20 times, but not remember a single thing from it. It’s drudgery and almost painful to read it. It’s a fight.”
His response was profound to me: “instead of you reading it how you are, try to understand why the author spent their life, time, and effort to learn that material and then convey it to you. What made them fascinated in it?”
This comment gave me pause… because it’s novel to me, while seemingly adjacent to my current worldview in a way that makes me suspect it’s an important missing puzzle piece. That makes me nervous.. yet excited.
I’ll have to meditate on this a bit more. Thanks for sharing!
too big of tasks. break them down and then proceed from there. have it build out task lists in a TASKS.md. review those tasks. do you agree? no? work with it to refine. implement one by one. have it add the tests. refactor after awhile as {{model}} doesn't like to do utility functions a lot. right now, about +50k lines in to a project that's vibecoded. i sit back and direct and it plays.
Imagine the CS 100 class where they ask you to make a PB&J. saying for it to make it, there's a lot of steps, but determine known the steps. implement each step. progress.
I was part of a shop that did the Pivotal Way and we had Inceptions where the PM, engineers, and a tester or two would be sequestered in a conference room for the day to bang out task lists that went into mid-level fidelity. Technical considerations were debated and sometimes in a heated way, but we never got into implementation—just structure and flow to ensure it jives.
I'm inclined to agree with this approach because someone not using AI who fails would likely fail for the same reasons. If you can't logically distill a problem into parts you can't obtain a solution.
I think you'll have a few more iterations until you figure out that going with this visual graph style isn't optimal.
reply