tyzoid

Why qwen 2.5 everywhere? Why not 3.5?

How does this work on pages that require JavaScript in order to render?

I'll sometimes use agent swarms for triaging quickly through a number of possible issues. For example, a single request for a code review to uncover all possible issues, followed by an agent swarm to dig deep into each issue to confirm or mark false positive.

I have to disagree with the fundamental premise here. LLMs do indeed have an internal "concept" mapping - the embedding layer. There's some fun demos on youtube[1] where you can manipulate embedding vectors (such as king - man + woman = queen), and these manipulations are done entirely in the embedding/concept space.

[1] https://youtu.be/gQddtTdmG_8 (around 14:00)

Domo Says No Mo' to AI FOMO

Hey, fancy seeing you up here on HN! Fun article

You can use something like flyway on top of your existing git/cicd stack. Write the query as a migration, have it reviewed using your git code review process, and merge to run the migration.

It doesn't even need to be a third party. Documents you make for yourself but never send to a third party could still be seized in a criminal context or subject to discovery in a civil one.

Attorney-client privilege is a special carve out because the courts have recognized that clients need to be open and honest with their attorneys to get proper counsel and representation.

This ruling is the court declining to extend that special carve out to non-lawyer AI tooling, and keeping the status quo of contemporaneous documents made by someone discoverable, whether or not shared with a third party. The judge draws from the TOS as an admonishment, effectively saying (my words, not the judge) the TOS should have put you on notice that you have no expectation this data is confidential.

Pretty much any legacy system can have a modern reverse proxy in front of it. If the legacy application can't handler certs sanely, use the reverse proxy for terminating TLS.

Upper-end estimates for typical web traffic CPC/CPM rates are around 10¢ per click and 0.5¢ per impression. I never click on ads, but let's hypothesize that with very-well targeted ads 5% of people click through. That ends up at around an estimated 1¢ per page view. Even at 5x that rate, I'd be able to read 1000 articles ($5 USD) before even approaching the $6 USD/mo rate I think is reasonable.

That's over 30 articles per day, again at a 5x rate than advertising will return. Will there users that read vastly more than that? Sure. But there's also many readers that will under-utilize the service too.

Just take a look at how YouTube Premium is doing, many creators report that their premium revenue vastly outpaces ad-supported viewers on a per-view basis.

If the revenue doesn't make sense, then you could supplement the revenue with ads for users who exceed a soft cap, or have tiered subscriptions. Something like a basic (1k articles per month)ad-supported subscription for $4, basic ad-free for $6, and unlimited ad-free for $10.

The problem is there's not really a good way to subscribe to these things. I'd gladly pay a nominal fee (~$6 USD/mo) for access to media, but I'm not about to subscribe individually to each site. Ideally, I'd subscribe to a single service and payment is split across the various sites in proportion to how many articles I read from each site.

There was a service that promised this a while back, but IIRC mozilla bought and killed it.

I've been doing this for most of my games too. I find I'm less likely to release a piece somewhere I don't want with this method.

Completely blank for me on mobile (javascript disabled)

My usual go-to for a quick static server is:

python -m http.server

But variations exist for a lot of languages. Php has one built-in too

who cares what experience in your resume was a lie and what not

Just being blunt: that's called Fraud. Making false representations for personal gain (employment, in this case) is one of the classic examples.

It doesn't matter if nobody checks in the moment, or if you usually get away with it, dishonesty is dishonesty. If I were to discover that someone joined my team under false pretenses, you can bet I'll have very little faith in their credibility going forward.

https://www.justice.gov/archives/jm/criminal-resource-manual... :

The Fourth Circuit, reviewing a conviction under 18 U.S.C. § 2314, also noted that "fraud is a broad term, which includes false representations, dishonesty and deceit." See United States v. Grainger, 701 F.2d 308, 311 (4th Cir. 1983), cert. denied, 461 U.S. 947 (1983).

The punishment needs to be commisserate with the crime, and dealt with through due process; to do otherwise is distinctly un-american (see 1st amendment on freedom of assembly, 4th amendment on freedoms from unreasonable state actions: "The right of the people to be secure in their persons[...] against unreasonable searches and seizures", and especially the 5th amendment: "[no person shall] be deprived of life, liberty, or property, without due process of law").

What we are seeing at Colombia University (as well as the country at large) is the continual abridgement of these rights. Note that for the fifth amendendment specifically, the constitution refers to any person, not just citizens. Those here legally are entitled to due process protections under the law.

The following argument relies on the following: (1) Universities historically have been the the catalysts of change through student protest. (2) Peaceful protest is a right of the people that shall not be abridged. (3) Public Universities (being government institutuions themselves; see campus police and jurisdiction) have a duty of care to protect their students.

With the above holding true, the argument against this being a "betrayal" falls facially flat, as it is a severe consequence that the university capitulated to, and had a duty to prevent. The arument boiled down to "they were being disruptive, so we should get rid of them," because the betrayal amounted to the jailing and deportations (or attempted deportations, in some cases) for the "crime" of being nonviolently disruptive in a public place.

Without articulating a legally rationed basis for a criminally sanctionable offense, an equivalent is threatening to jail and deport construction workers when they block a business entryway. In general, you do not have a right to be merely inconvienced by others in a public space.

1M revenue isn't that high a bar to clear in retail, just takes one popular/meme product. After all the COGS/fixed costs are tallied up, that could leave you with significantly less with which to contemplate custom development or platform changes.

I'm not seeing them show up, with or without JS enabled (firefox on android). I might suggest having some interaction for non-js users though (details element, perhaps?)

Most likely, yes. Then it wouldn't have mattered that the `sl` package was installed.

Sometimes a KV datastore is the right abstraction, though. Caching is an excellent example, but also for distributed session storage, configuration management, nonce enforcement, etc.

I'm curious, did he not clean out old chats that were no longer needed? If UK FOI applies to this now, does that introduce recordkeeping/retention requirements?

I'm not a lawyer, but the legal claim made appears to me to be on shaky ground. In my understanding, there has to be actual damages arising out of an action. "I could have been hacked, so I had to spend time/money on it" isn't actual damages unless they were _actually_ hacked.

You can't easily take text and make it red that way, only edit the existing content.

If you're not the copyright owner, that's a terrible idea. DMCA notifications are submitted under penalty of perjury:

17 USC 512c3(vi) A statement that the information in the notification is accurate, and under penalty of perjury, that the complaining party is authorized to act on behalf of the owner of an exclusive right that is allegedly infringed.

Cue Charlie Munger's "Just replace EBITDA with 'Bullshit Earnings'" https://youtu.be/7B_6AFG0lUU

It's a chromium derivative.

I've found them to be quite accurate when given enough context data. For ex, feeding it an article into it's context window and asking questions about it. Relying on the LLM's internal trained knowledge state seems to be less reliable.

Why should they? If they've got a workable solution and are living a happy life, there's no reason to treat it as a problem.

True, but it dissuades folks from using directional signals, broadcasting RF energy in more directions and increasing the noise floor for everyone. I feel like there should be some sort of middle point here.

Discord recognizes <t:{{unix_time}}:format>

https://discordtimestamp.com/