Worth considering how car-centric America is. An 8-year-old is unlikely to have access to a grocery store to which they can independently travel. Once they're at the store with a parent, they'll just travel the aisles together. It's not as though many young children have the funds to make purchasing decisions, anyway.
This has been ubiquitous for a while. Even here on HN every thread about these models (even this one, I'm sure) features an inordinate amount of disagreement between people vehemently declaring one model more useful than another. There truly seems to be no objective measurement of quality that can discern the difference between frontier models.
Running a model comparable to o3 on a 24GB Mac Mini is absolutely wild. Seems like yesterday the idea of running frontier (at the time) models locally or on a mobile device was 5+ years out. At this rate, we'll be running such models in the next phone cycle.
Seems to be a significant number of people who have deemed LLM responses 'good enough' and completely dropped search engines altogether. I would imagine that works fine for people whose queries are simple and/or the accuracy of the result is not actually important. We may be discovering many people just wanted Google to tell them what they want to hear and LLMs are much better at that than scanning a handful of garbage Quora posts.
The web is going to need a different business model. There's no way around it from this point. Traffic + Ads = Money is dead for anybody with content that can be summarized by an LLM. Even Traffic + Subscription = Money is also vulnerable if your offering is not easily discoverable outside of a search engine. I don't know the solution.
This one, specifically, drives me crazy. The boldly colored 'Continue' does NOT, in fact, continue the results. Instead, it continues to the app store. Just awful.
We've acquired a handful of their books and they're just not great for reading to a toddler. The dialog is in speech bubbles, and the layout is more like an episode. I have to come up with filler text and structure on the fly. Give me a Franklin or Little Critter, any day.
That episode perfectly describes my experience as a father of a toddler. I'll decide on a whim to taking him hiking only to discover half-way through that I didn't bring enough snacks for him and forgot his water shoes at home. So now I'm out with a cranky toddler who's hungry and can't play in the river. I've learned to accept my wife's 30-minute packing phase to send us on our way with everything we need.
Bluey is the best children's cartoon of this generation. The art is a joy to see, being full of color and well-defined lines/objects, without being overwhelming. I sincerely appreciate the creator's resistance to engineer the show for maximum attention and, instead, focus on telling a good story with relatable characters. The show can move slowly when appropriate and isn't afraid to keep the scene mostly static. The music is great and well directed to capture meaningful emotional moments without being campy. The episode length of ~7 minutes is perfect, especially for families wanting to limit screen time, as it covers a quick breakfast/lunch (or laundry/dishes break) and provides a natural stopping point to move to the next activity. I can't praise it enough and just hope its success inspires more studios to avoid the brain-rot algorithmic-lead nonsense.
My only complaint is that I am a grown-ass man and shouldn't be tearing up every other episode of a children's cartoon...
Gambling and other predatory businesses on military bases are gross. MLM's run rampant through military communities by preying on dependents and the tighter social networks formed by military families. Yeah, I know, freedom to make adult decisions and all, but sometimes it's nice to just keep the garbage out in the first place.
> In the test, a human evaluator judges a text transcript of a natural-language conversation between a human and a machine. The evaluator tries to identify the machine, and the machine passes if the evaluator cannot reliably tell them apart. The results would not depend on the machine's ability to answer questions correctly, only on how closely its answers resembled those of a human.
Based on this, I would agree with the OP in many contexts. So, yeah, 'basically', is a load bearing word here but seems reasonably correct in the context of distinguishing human vs bot in any scalable and automated way.
I'm really waiting for somebody to figure out the correct interface for all this. For example, study mode will present you with a wall of text containing information, examples, and questions. There's no great way to associate your answers with specific questions. The chat interface just isn't good for this sort of interaction. ChatGPT really needs to build its own canvas/artifact interface wherein questions/responses are tied together. It's clear, at this point, that we're doing way too much with a UI that isn't designed for more than a simple conversation.
The amount of information the SVG must store to represent the animation actually crashed the application creating it. There’s too much going on in TTE animations.
The workflow only takes a few seconds. The .cast file is output to a folder, AGG reads the .cast and creates a gif. AGG has arguments to automatically set the row/column widths and strip inactive time.
I could not find any other solution that produced high quality gifs. Asciinema is awesome.
I'm sure somebody is going to point out that it was written by AI and is a toy, therefore it can be maintained by AI. I share your desire to have human maintainable code, but I imagine one of the goals of AI written code is to allow the AI to manage it, end-to-end.
I have found myself leaning heavily on easing functions to smooth motion within my terminal visual effects engine. I do not have a very strong math background so I am limited in how much I can modify the commonly available functions. I have found it useful to create custom easing functions by mapping the easing function progress across a bezier curve. There's an example in the changeblog write-up from the last release:
I came to the comments to see if anybody else was bothered by this style. Every other paragraph contains a list rephrasing a simple concept. Bit of irony for an article focused on the problem of compression to so easily exemplify where compression would be useful.
News websites, especially local news, seem to have completely missed the meeting on how to build user friendly websites. They are almost universally full of obtrusive advertising, obtuse layouts, irrelevant links, and poorly handled media players. They are really in a league of their own for bad user experience.
Then don't use the service? Use one of the 'better' free alternatives. Let's see how long they remain free once (if) they actually see a meaningful amount of traffic. Complaining about advertising while using a service for which you pay nothing is immature and unreasonable. I hate advertising and have done everything I can to remove it from my life. That means paying for services that I enjoy using. I would rather provide value to the business via money than ad attention.
Well, this is cool. I'll have to see how it handles the sorts of effects I show in the README at https://github.com/ChrisBuilds/terminaltexteffects. I don't know much about SVG but anything that attempts to actually store the text is going to create a very large amount of data. I'll try it for fun.
And somehow actually paying for youtube which fully removes advertising and provides revenue to the service/creators is seen as an utterly absurd proposition by a staggering number of people.
There has been reporting about this for months. I realize I'm just a pleb, but I really don't understand why a company with, essentially, limitless money, won't simply install the emissions reducing hardware and properly permit the turbines. Like a lot of what's going on around Elon/US politics, it seems like a meaningless fight and is only further torching reputations. Just do the obvious right thing and move on.