cyanregiment

Everyone about to update all their WoW addons

I'm from nor cal but always liked Mistral.

Mistral 7b is still one of the best free/open models you can run locally on a MacBook. So fast too.

Looks sick.

Should leverage the quick access of the browser by making the demo account-free so you get right in.

Games with launchers can't compete with that.

• 996 schedule

• Ads all over SF that just have our logo with no context

• Hacker house in Pac Heights

• Shared space in Hayes (for calls)

• Tech stack is Next.js/React/OPENAI_API_KEY

• Product is AI SOC-2 verification

• The company will fail, I'll pivot into real estate

Idea is great, but execution is awful - sounds like a screen reader.

I'm waiting for the text-to-Joe Rogan Experience model.

I should at least get equity on each login.

After 1000 logins I get the source code.

This is too much like rent, it's not fair.

Something something, cryptography.

Come on crypto bros - we need you!

The internet money thing aint gonna work. Look at the pyramid on the back of the dollar. It's esoteric. They will make you disappear and make it look like an accident. Look what happened to Ross.

This isn't even a reply anymore, sorry knollimar

It's like a cartoon where the monkey finds the ruby, another takes it from his hand, another takes it from his hand, then a gorilla bonks them on the head and runs off with it, only to have it snatched by a swinging monkey who loses it to a python who drops it into the mouth of an alligator... who brings it deep underwater... accidentally swimming into the mouth of a giant whale, where inside the whale is an island where he coughs up the ruby, it's ejected out the blowhole high into the air before landing atop a tree and hitting every branch on the way down landing before a monkey who finds the ruby, another takes it from his hand, another takes it from his hand...

Was this one the last stral for you?

The stral the broke the camel's back?

When I mow lawns I teleport forward 6ft at a time and then a 6x6 square of grass vanishes beneath me.

In the nude.

I'd say the inverse - it's only not true when it's a bad UI (like ClickUp's).

I agree though, sometimes text input is a superior UX than using any UI - like Spotlight search or git CLI.

Also agree that AI will broaden that - but the default?

A good thing to keep an eye on is image editing: Will people keep telling a chatbot what changes to make, like a slot machine hoping for the best, or will companies start putting more tools into the workspace (e.g. draw a selection box around an area and "Generative replace...")

People need to stop pretending text and voice commands are an evolution on the user interface.

The UI is the innovative thing that simplifies and enables focus - not talking to “an agent”.

Imagine saying or typing what VIN you want instead of a dropdown menu with a list of them - isn’t it way faster and easier to see options and click one in a UI? Rather than have a back and forth conversation?

The take is really - if you zoom out - that we can cram all UX into a chat. Idk about that, I like a UI.

I’m more interested to see what kinds of UI/UX beyond chat we get. If you’ve messed with the AI video SaaS that’s already out there a few UX paradigms are already being established, like the concept of image references - they all do it different ways, but there will be an established way. Like pull to refresh or pinch zoom became established UX for smartphones.

You need to be online to prove you own it, which now means you don’t?

Part of the overall trend to eliminate in-person shopping and physical product ownership.

In the end, 1 guy owns everything and we’re all batteries.

Accurately? About a ~500 line React component or API endpoint given lots of very good context

Which is not bad

I don’t bother with the agentic stuff anymore than I would run a calculator on auto - what is the actual point if you think of what that really means

You don’t need React until you do.

And that’s fine. It’s easy to replatform later with LLMs.

With vanilla you typically do class based components for organization (which can be quite clean in type=“module” with import/export) and you end up writing a complex View type class that does all the DOM manipulation.

You gotta hide the framework somewhere or you’ll repeat yourself a lot

Bonus points at scale you invent a templating solution. Maybe you use JSX

All the expensive planes are falling out of the sky too. It seems like some weird plan to unwind the US ever since Obama left office.

There’s no growth or hope. It’s all bad news, war, and rumors of war - and all reads like a bad WWE script.

Just seems like a coordinated effort to destroy from within since about ~2016 or at least act out this play of America in Decline

Whatever, would be way better if they celebrate collaboration and forward thinking again.

Mainly because it’s just so boring under Trump - this should have been a 4 year show

Yeah, but because it's an LLM it makes very good guesses. It's not so random, it's seen URLs in training data.

One of the points of the experiment was to see how much of a shortcut using an LLM could give me, as the developer.

I've done other LLM search experiments where I basically just prompt to "predict 25 search results for this query" and it's surprisingly good. You can also run a check to verify the URLs to splice hallucinations out of the result, but yeah, LLM is already doing a lot of the work for search.

It's estimating one's own ability at performing a task - not "confidence" in a personality sense.

Someone can estimate their own ability at something at 0 and because of their own confidence and determination, they - to your point - may outperform a less confident person who also estimates their ability at 0.

The point of the study was that those who estimate their ability higher generally performed worse (regardless of confidence, impulsiveness, which we more broadly attribute to general "confidence" in personalities).

It makes a lot more sense if you think of it as a mere “effect” and not a “syndrome” the way it’s used.

Simply: Anytime anyone overestimates their ability, they will perform worse than if they had approached it with humility (on average).

Rather than speaking it like people “have Dunning Kruger”

Just a fact (heuristic) of behavior that anyone can fall subject to

If you’re talking about SSL cert logs, that excludes a ton of web content.

And you haven’t shown me that you can even get all those for free anyway

Looks like they only give you 100 on a trial account

That's what Rust looks like? Coming from JavaScript it's surprisingly clean and readable. Maybe I'll actually try it out.

Is there another way to install your HN CLI?

The exhaust from a hydrogen fuel cell car is drinkable water.

However, experts do not recommend drinking it straight from the tailpipe because the water can pick up dust, dirt, and chemical residues.

(it had to be disclaimed!)

People are critiquing whether or not AI wrote the article.

Well, it's clearly not 1-shot / low effort.

(can we agree there's a difference between low effort 1-shotting a prompt and those who use LLMs interactively?)

The one reason why I can't take a mega hardline stance against AI for writing: Not everyone speaks English as a first language.

People with awful English speaking and writing skills are using LLMs to contribute content in English. This is a good thing. They used to try their best and suffix every post with "sorry for my english".

The author's name is Ettore Di Giacinto - I put 2 and 2 together on this one.

I give a pass when non-native English speakers are just using it to write better English and keep up with the rest of us. It wasn't super obvious anyway, it wasn't interruptive.

The content is beyond that, and I more than welcome stuff like this here.

502 - Bad Gateway

Another cool thing about Mistral 7b is it runs offline, and I don't need to download any massive URL list that may or may not be current.

It does the work of finding URLs at runtime, offline (until it verifies the URL exists).

The author asks more than a few times:

how can we know ___ by only examining language

I've accepted that language is knowledge. In the case of LLMs, to model language is the same thing as modeling knowledge.

Webster's: (1) information, understanding, or skill that you get from experience or education

If we look at knowledge separately from experience and education - as just the raw knowing - what we mean exactly by "having knowledge" is that the language therein is of a certain (correct) kind and order.

If I can reliably say all the right words, in the right order, about X, then I know X.

My crawling approach did hit Wikipedia several times, and crawled a number of pages, but not the entire site at once.

With either approach, we would both eventually end up with all the websites - that being a surprisingly small number (~1 billion of which we only ever access ~1000).

I got in an adult content bubble from starting with "A" ("Adult") and would have hit another one presumably at "X". But "B" had quite a few too, lol.

Most of the words were bubbles - where it spent a decent amount of time on each one and I wondered if it would ever move on or was stuck. There are just a lot of backlinks and tangents for every word you can think of.

Another fun thing about this experiment is you will find a lot more international websites that no other search engine will find - some weird personal sites and some surprisingly good forums.

But yeah, you could enter anywhere - Wikipedia, or some other random Array of words. It doesn't have to be a dictionary.

Don't do that, you might import a std