uses

How does this work? Does it feed the PDF's text as a prompt to the LLM? How would you do this if you had, say, thousands of pages of a website?

I feel like "chatbot/search engine hybrid which can consume a large website and know everything about the org it represents" is a powerful application.

How close are we to handing this thing a desktop and an internet connection with the prompt "ok now make gpt-5"?

In fact, the models appear to be already kind of doing that? With the fuzzy layer of the humans still in the loop.

I was just hearing about a local who's using ChatGPT to write copy for their personal website. Then I was thinking, how long until the browser itself has a built in feature to summarize paragraphs of text (which was possibly generated by a model) into chunked, readable content, replacing the original html as the page renders? These models are consuming the world and each other faster than we can think of what's next.

I mean you could just stand in front of a self-driving car and it'll probably stop rather than swerve around you, I imagine. There was an episode of revisionist history that discussed this a bit https://www.pushkin.fm/podcasts/revisionist-history/i-love-y...

Ah, more journalist hate. Let's blame the journalist for writing a well-researched, thorough article citing multiple primary sources, because it tells us facts we don't like. It's gas lighting by the elites after all! This is why, whenever I read an article about anything, I blame the gas lighting elite journalists in their cozy chairs with their opinions backed by nothing. Those damned journalists, always reporting things.

What? Those are exactly the type of solutions economists would prescribe.

I was just thinking about how much world-class content is available between podcasts and email newsletters, and it's all free. Whereas if you wanted to say, read an article on the web from the same sources, you're looking at a huge paywall. The "podcast app" market is heavily saturated with many great competitors. Yet the "newsletter reader app" just...doesn't exist? I'm stuck using iOS Mail with its lack of reading options, no saving my place or otherwise bookmarking. I just can't believe nobody's made something to fill in this need?

Personally I’d like to hover over the weather widget without seeing tabloid and politics headlines. There doesn’t seem to be any way to remove them.

HTML / CSS ended up being extremely granular and hackable while missing the big building blocks that would have readily matched the structure of the web - things like navigation menus or page outlines. We waited over a decade for the layout systems to catch up to developer needs while creating buggy float-based grids. It's an outlier that media queries were already in widespread availability when mobile really started picking up.

This episode of an excellent podcast about Xi covers the Covid response really well

https://www.economist.com/podcasts/2022/09/28/8-the-great-he...

How am I supposed to match two entirely different words? Atomic groups with | don't seem to work. Not something I'd use regex for anyway. What am I missing here?

Wow, this is still going on? I remember it being the case for Windows laptops and pre-builts ~15-20 years ago.

I've been using self-built desktop PCs and Apple laptops for so long I've become out of touch. I'm - somehow - surprised this is still going on.

Pretty surprised they left the front camera on the side of the device. I can't figure out why they would think that makes sense after using it even once. It's so awkward trying to do a meeting and I have this weird camera angle coming from the corner of my face. The alternative is portrait orientation, which puts the camera really far from the center of the screen - feels like it's coming from above or below my face - and puts my video feed opposite to the orientation of my audience's screens, while also not being able to lean on the folded case.

One is a core strategic element of a trillion dollar company, the other is donationware?

In any case Chrome's dominance is more about Microsoft's failings than Mozilla's.

How would a USD CBDC interact with Fednow? My understanding is that Fednow will provide a way for fast interbank transactions with instant settlement because the accounts will be held directly at the Fed, by commercial financial institutions, on behalf of consumers. Rather than the current ACH system where banks settle between each other after several days via the clearing house.

If I have deposits in US CBDC, who will act as the equivalent of my credit card company or credit union? For example, who protects against fraud or reverses transactions? What would be the benefits of having CBDC deposits vs having a Fednow-enabled account at my credit union?

It’s weird you didn’t mention the number one reason for all that. Millions of people with decent incomes and nothing to spend it on due to Covid, bored out of their minds and looking for something to do. Unfortunately for them they did something really stupid. At least nobody got hurt.

Is anyone looking into what it means when we can generate infinite amounts of human-like work without effort or cost?

> Curbing Misuse [...]

That's great, nowadays the big AI is controlled by mostly benevolent entities. How about when someone real nasty gets a hold of it? In a decade the models anyone can download will make today's GPT-3 etc look like pong right?

Recommender systems etc are already shaping society and culture with all kinds of unintended effects. What happens when mindless optimizing models start generating the content itself?

The weirdest thing about the Advantage is the hollow echo sound it makes. I'm using a $300+ keyboard but it sounds so cheap due to the hollow sound. It seems so obvious after first typing on it.

Besides that, it's an excellent keyboard and I'll never, ever buy a regular keyboard again. Everything else about the design makes sense, and it's completely programmable.

This reminds me of the chat situation where a coworker will start with "Hi!" typing indicator.

In either case, it's like "yes, you have my attention. I cannot focus on anything at all until you hit enter, and I also can't reply because I have nothing to reply to".

I also can't complain like "hey if you're going to interrupt me can you just say the whole message so I can actually reply instead of being held in absolute suspense for many seconds of my life?" because that would be, frankly, even worse than what they're doing, which was done out of a kind of ignorance of etiquette.

Maybe I'm too brainwashed by git, but it seems to me one of its benefits is the way the index works. You're encouraged to be fairly explicit about what you're adding to a commit, which encourages making a nice version history. Why would I want everything I do to automatically be added to a commit by default? Doesn't this encourage me to either put all kinds of unintended, not-ready crap in my commits, or constantly manage the equivalent of .gitignore? What am I misunderstanding about the benefits of how this works?

What's a good next step after learning "classical" neural nets, i.e. backpropagation ANNs?

I've been working with MNIST with C and CUDA with dynamic parallelism for a couple months and it's been extremely enlightening but I'm kind of ready to move on.

CNNs maybe?

Inflation doesn't exist as a policy to stiff wage earners. Inflation exists as a reality of a growing economy (more goods and services needs more money to cover them). The alternative is deflation, which would be extremely bad. So what we aim for, is low inflation. Which we've had for ~40 years. Low, controlled inflation is the vastly preferable alternative to any amount of deflation, so it's the side of the scale we target, giving a ~2% margin of error (~2% is the inflation level usually targeted by modern central banks).

What Krugman is talking about is that cutting wages is extremely hard for employers to do. Nobody wants to take a paycut. So what employers do instead, being put in the situation of needing to cut payroll, is eliminate entire positions. Which on the whole is worse than a paycut because of the destabilizing effect to the individual and society. So a positive side effect of inflation (not the main effect, a side effect), is that in the absence of pay raises, low inflation creates a constant mild paycut unless there is specifically a pay raise. It makes the economy more resilient.

What did they end up doing with the escape key?

So, do I understand correctly that these are the main benefits of a CBDC?

  - A centralized ledger which is maintained by the central bank in question.  
  - Instant payment clearing - no ACH or other steps. You simply request a transaction and it happens immediately if it's possible.  
  - No fees - presumably you can send large amounts without paying for a wire transfer.  
  - Convertible into "regular money" with no fees or intermediaries. You can exchange this currency for traditional deposits at a regular bank, and vice-versa.  
  - Easy transactions with the government. You can pay your taxes, or get paid (stimulus payments, paychecks, for example), instantly.  
  - From the central bank / policymaker perspective, there are many benefits because of the gigantic amount of data they gain access to.

Would this be essentially like Fedwire, but for everyone rather than a special service for banks?

That table is a design fail... the 4th column should be the 4th row, like this: https://imgur.com/a/mTRONdV

Discord is amazing for getting fast answers to questions, but terrible at accumulating knowledge.

For example, several times I've joined a server to ask a question about an obscure game mechanic, and gotten helpful answers in a few minutes.

The flip side is none of that information is crawlable on the web, so it is lost like sand through an hourglass. Whereas with forums (or their replacement, reddit) somebody in the future could benefit from my Q/A session by finding it on a search engine.

The second major problem I have is identity. When I post in a channel about school, I want to use my real face and name. When I post in a random server, I want anonymity.

The processing doesn't really bother me much, because AFAIK, processing itself has no relevance to health. The problems with modern postindustrial diets are that calories are too cheap, available, and delicious. I think the best long term solution to this will be some kind of food science that adds "filler" to delicious foods that simply passes through the system and adds bulk without decreasing deliciousness or adding excessive calories. All the attention to healthy lifestyles is great but it will continue to only be practiced by the minority of people with the resources (mental, cultural, monetary, etc) to spend on those considerations. Meanwhile obesity, heart disease, etc is a major drag on our collective health and QOL.

> Remember when the previous administration suggested looking into the Lab Leak theory and was laughed at?

The previous administration was completely nuts and would do and say all kinds of random things. The problems with communication and trust were too many to know where to start. Besides that, the context was totally different. A president trying to "wage a trade war" on China, stir up nationalist sentiment, distract from his responsibility to respond to the pandemic, constantly regurgitating conspiracy theories and other misinformation, constantly trying to distract from scandals, etc.

> Remember how you would get banned from Facebook for suggesting that the lab leak theory was a possibility?

The amount of misinformation about COVID was and is extreme. Yeah, they moderated their platform, which is under intense congressional scrutiny, in the middle of a once-in-a-century pandemic. Even so, I think you're conflating the bio-weapon theory with the lab-leak theory.

> Then major social networks and the medias reversed their positions. What changed in the meantime, other than the president?

Major social networks and medias didn't "reverse their positions"...we learned more, we're still learning more. What changed in the meantime is so, so many things lmao.

Bloomberg Podcasts did a 7-part deep dive into WeWork last year and its extremely entertaining and eye-opening: https://www.bloomberg.com/news/articles/2020-06-25/wework-po...

I believe Ellen Huet was the narrator and main journalist. One of the more fascinating podcast series I've listened to.

Hearing about Adam and all the wackiness surrounding him, I can totally recall some of the hyper charismatic people I've known. They can have this weird almost energy field around them that just turns people's brains off.

Is this some kind of advanced google PR algorithm move to finally kill off something that nobody will mind, because nobody knows it exists?