voiper1

It seems many of these - e.g. turbotax - are more specifically regulatory capture. The capitalistic business that makes a profit on tax preparation surely opposses the government giving it away for free.

Is there a state of the art for face matching? I love being able to put in a name and find all the photos they are in.

I don't even mind some training of "are these the same or not"

That's one of the conveniences that means I'm still using google photos...

Nope - there's at least one layer of safety:

>For security, all template variables are escaped:

>// This will run `ls 'foo.js; rm -rf /'` >const results = await $`ls ${filename}`; >console.log(results.stderr.toString()); // ls: cannot access 'foo.js; rm -rf /': No such file or directory

>But no one is asking, "What will we do if" – when – "the AI bubble pops and most of this stuff disappears overnight?"

We'll just run self-hosted models, they're already becoming reasonable.

A font-viewer for a specific phrase, with my own chosen list of fonts. Mostly for seeing how hebrew looks in different fonts for family name signs.

https://avimar.github.io/fonts-hebrew/demo.html?text=%D7%97%...

This is definitely what I want... forget privacy, the world is all in on convenience. Let GPT4 look over my shoulder at my screen and also listen to my voice and automagically tell me useful info. Of course, I should be able to copy/paste or otherwise insert the output.

I didn't see an example and I can't quickly think of a prompt. Did it do "better" by itself, or did telling it it's important trigger chain of thought where instead of flippantly giving an answer, it said all the steps explicitly? We know Chain of Thought can definitely improve performance.

That's not a full answer, it's still really cool that "this is important" would trigger a more careful analysis, especially since it doesn't "know" chain of thought produces better results, but that's a far different implication than "say the magic word" gets better results.

Whoever editorialized the title didn't have much in the article that actually said that...

I'd assume that people why are handicapped in some way may be better able to manage without a commute while working from the comfort of their home environment that the have control over.

Yes, a programming job may only require sitting at a computer... but also commuting to the office. WFH means you don't need to deal with the commute.

NOTE: title doesn't match the article title.

Article title: Does working from home damage productivity? Just look at the data.

HN title: WFH significantly increased workforce participation from those with disabilities

The first comment is from the Cofounder and he lists the _features_ that motivated a rewrite:

>Then there's the question of why we did a rewrite at all. We wanted to get at some important requirements: >Unlimited cardinality >Analytics queries against time series at the performance of a columnar DB >Use object store as the durability layer for historical data (i.e. separate compute from storage) >SQL and broader ecosystem compatibility >All of that stuff taken together meant that we'd be rewriting most of the core of the database. ...

I don't buy this cynical take. The current water treatment plants are much more complicated than a simple septic system which lack economies of scale and can still lead to ground water pollution.

Also, we've apparently transitioned from waste water being something that had to be dealt with to an actual asset. It can be cleaned to partial usage (and used for agriculture) or directed to help refill specific aquifers.

They want it to return a single token yes/no, which may not work so well since it doesn't have "space to think". Chain of thought is much more reliable.

But that costs more.. but they ended up anyway doing: >The other key will be 'reason' and include a free text explanation of why you chose Yes or No.

But they did yes/no FIRST, then reason. So he ended up asking for the answer, and then asked it to _justify_ why that's the answer. For chain of thought to be helpful, you do the opposite: First explain why these addresses match or don't match, then give a final answer. Same amount of tokens but activated chain of thought prior to the answer, giving it "space to think".

Only the creator needs to have an account, not all users. If you like jitsi or don't want google, etc, then just start it and nobody else needs an account.

You send your AI with what you want to share, they send their AI with what they want to share... google figures it all out and we just get emails with the input. Perfect!

(2021) I thought the picture looked familiar.

When searching, I find everyone referencing the two-sleep information - much based on Roger Ekirch.

https://www.sciencealert.com/humans-used-to-sleep-in-two-shi... https://www.discovermagazine.com/health/dont-worry-if-you-ca...

However, I did find this article from The Atlantic: https://archive.is/MYcgm#selection-1033.0-1049.105

>But humans have never had a universal method of slumber. A 2015 study of hunter-gatherer societies in Tanzania, Namibia, and Bolivia found that most foragers enjoyed one long sleep. Two years later, another study found that a rural society in Madagascar practiced segmented sleep. Two years after that, a study found that the indigenous residents of Tanna, in the South Pacific, largely had one uninterrupted sleep.

I write emails and put it into chatgpt and ask it to make it more concise or point out issues. No utility in asking chatgpt to needlessly expand the text...

Yep, copy/paste issue on desktop is an issue.

It's even worse on mobile where I'll often have to send myself a link (sometimes one step back to how I got there) to deal with it on desktop.

You can also try Whisper, or try google or amazon's ASR and give it hints/dictionary of common words.

They have their own ASR Conformer-2[0] and support 9 languages (they count it as 12)[1]

It looks like their synchronous transcribe is much slower than whisper, but if you need it fast, you need their realtime ASR (or amazon or google's).

[0] Conformer-2 is trained on 1.1M hours of English https://www.assemblyai.com/blog/conformer-2/ [1] https://www.assemblyai.com/docs/Concepts/supported_languages

I use it much less than before -- but I still think it's very useful. I think it's because I have to go out of my way to use it. I still haven't figured out how exactly to make it a natural part of my workflow, since I don't need it for any one specific thing.

While you've clearly explained how this is bad, how is this not a free market?

>an economic system based on supply and demand with little or no government control

... or the hacking. They mention exploring the security implications.

OK, so this wave of LLMs doesn't eliminate blue collar or the service industry.

That leaves "just" removing white collar jobs, which is estimated at 43% of US jobs!!

https://www.dpeaflcio.org/factsheets/the-professional-and-te... (It seemingly includes nurses and teachers, which may need to be removed)

If you use ChatGPT as prescriptive, that it tells you what to do, then you're only as good as what you put in and how good ChatGPT is.

But if you use it as a coach/consultant/pair-programmer/R&D exploration/ brainstorming session, then you have instant access to an "expert" in any field.

That's something that can supercharge the productive output of any worker. And/or lead to dismissal of most of the team...

Incredible detective work!

Why would discord let anyone delete a webhook?

I'd think anyone can post to the webhook, but you need to be authorized to modify it.

Indeed, tracer bullet is immediately what I thought of from The Pragmatic Programmer.

I'm pretty sure "due process" means being tried in a court of law. Buying data doesn't skip being tried in court.

I assume you're implicitly referencing that the FBI is skipping getting a warrant, but why do they need a warrant? Your information is already publicly for sale.

As other's have mentioned, it's not really an FBI issue, it's a general privacy issue - that companies are collecting, compiling, and selling this information and it's all legal. Which is darned hard to get people to care about - because it's convenient to give your information away. I know about these things and still use google for almost everything (I finally moved my business email to fastmail... but didn't yet bother to move my personal email.)

And even if the new _is_ better, it's called the bleeding edge for a reason. I find I bleed all over it - stress, bugs, and time.

If it's not worth spending one of your innovation tokens[0] on this new fangled technology, then use the old version that doesn't require the bleeding.

[0] https://mcfunley.com/choose-boring-technology

I don't get what the big deal is about the name.

Yes, I like linode and run most of my business on it.

They changed the name.

I'm not switching until they "kill" the product. If the uptime remains good, the service remains good, why switch??