Not a great metric, research in academia doesn't necessarily translate to value. In the US they've poached so many academics because of how much value they directly translate to.
This stuff is funny because it shows how worthless the actual service is on top of it and its mostly a wrapper in race to bottom. Same thing when cursor talks about how growth was flatlining until gpt-4 came out... So many of these companies need to disappear.
I feel like Apple makes this very hard and against their ToS right or in a gray area, aka shut it down at any moment. There are businesses that solely try to manage provisioning of verified apple accounts and having worked with a bunch of them reliability is pretty hard.
Having worked with mobile messaging and AI and trying to wrestle with iMessage and WhatsApp because I'm pretty sure this goes against ToS as they monitor how many new contacts and messages you're sending and get flagged to be shut down... If this is not the case, it would be very interesting....
Claude may write all the code but this is an oversight from the dev. Do people think these agents are acting independently? If they wanted or had thought of tests that would catch this then they would have them! The use or non use of LLM is irrelevant. I find the discourse around this all so strange.
On the other hand people ask "where is all the amazing software that has been vibe coded, I haven't seen it?". So Claude Code is two things at once (1) incredibly popular and innovative software that's loved by a huge amount of devs (2) vibe coded buggy crap. If you think this bug is the result of vibe coding, frankly you should look at Claude Code as a whole and be impressed with vibe coding. If Claude CLI has been "vibe coded" then vibe coding must be fine because I've been using Claude Code for probably 8 months and it's been a pretty smooth experience, and an incredibly valuable tool.
In what world do you get that conclusion? Dense cities in other parts of the world rely on mass transit to move people. FSD so you can have self driving cars in the street -> ? -> increasing congestion. The point is you should have more effective volume transit not optimizing random ones. 1000 cars on FSD are an optimization better than 1000 taxi drivers, compared to a train or a few buses.
> Dense cities in other parts of the world rely on mass transit to move people
But they don’t solve the fundamental problems with mass transit: comfort, convenience, and trip times.
I was just at Disney World and joking to my wife that the park is propaganda to get people to use public transit because Walt loved trains. (You can’t get into the parks by driving. Only the train, bus, or gondola. If you drive you have to park at this giant lot and take the train from there.)
But even within that self contained environment—where land for transit infrastructure is basically free because Disney bought it all decades ago—public transit still isn’t great. Trying to get three kids and all their shit out the door to catch a bus or train, having to fold up strollers, etc., just sucks in comparison to throwing them into a car.
Clarification: only Magic Kingdom has that extra step of train/ferry/bus transportation from the parking lot. For all the other parks at Disney World, you can drive in and walk to the entrance. (There are trams if you're far out, but you can also walk.)
Yes, Walt did want that extra step of separation from the real world. Then Disney realized the logistical problems with it and didn't repeat it for the other parks, but they're stuck with it at Magic Kingdom. (And it also acts as an upsell for the premium resorts that are positioned right next to Magic Kingdom in walking distance.)
Let me tell you something about Magic Kingdom: Everything between the parking lot and the rides exists the way it does in order to SLOW YOU DOWN so you spend more money. All of it! Okay, maybe not the parking lot trams, but all their other gimmicks are just part of a funnel to keep you away from the rides as long as possible and once you're inside, to keep you from leaving.
Would you even be in favor of a public transportation system that did what you’re asking for? If MAJOR_CITY had a bill for an Uber-like system where you could hail a car subsidized by the taxpayer or even free to give you a ride?
Or are you against the idea of “public transportation” (transportation costs shared by taxpayers) and your criticisms of its quality or convenience are really moot because you are ideologically against the premise?
Quite the opposite. I empathize with people who have to yell at their kids to get ready so they can make a fixed bus or train schedule. Or have to walk with their kids to a transit station in the rain. Or have to sit with their kids in a train while someone freaks out next to them.
Transit advocates are the ones who lack empathy for the typical person whose life is improved by having personal, point-to-point transit.
In the thread above, someone proposed cities investment money in FSD, and then someone responded: “Dense cities in other parts of the world rely on mass transit to move people.” So the discussion here is about how the government should prioritize investment in technologies.
Someone using a train doesn’t mean someone else can’t use a car, but when cities make decisions about their transportation strategies, they have to prioritize.
How selfish does someone have to be to see dozens of examples of world class cities with trains and roads, then think that trains shouldn't exist and people without cars shouldn't be able to get around just in case roads could be improved in some secret undetermined way.
Even the current amount of buses with some automated routing and phone booking would be a tremendous improvement. (Sure, there are accessibility concerns to address)
lots of highfalutin language trying to make something thats pretty hand wavy look like it's not. Where are the benchmarks? The "vector algebra" framing with @X + @Y - @Z is a falsehood. Embedding spaces don't form any meaningful algebraic structure (ring, field, etc.) over semantic concepts, you're just getting lucky by residual effects.
I'm giving you, the user, the easiest ability you've most likely ever had to explore embedding space yourself. Embeddings are tricky and can mislead, but they do often compose surprisingly intuitively, especially when you've played and built up a bit of an intuition for it.
What is the impact of misleading embeddings, how do they compose? I honestly am interested but don't know enough to understand what you're saying.
Why would I want to explore the embedding space myself, isn't this a tool where I can run cross-data exploratory analyses against unstructured data, where it's pre-populated with content?
We can iterate fast with understanding useful paradigms of vector manipulation. Yesterday I added `debias_vector(axis, topic)` and l2_normalization guidance.
The manifold structure of embedding spaces isn't semantically uniform, you've found a nice little novelty thing but it's not rigorous, and using AI slop to name this vector algebra instead of finding or running a benchmark to show that its actually works better.
I think a missing key detail that becomes obvious in hindsight is that you needed a lot more structured logic before when llms were not as powerful, but now that these models are being post trained on this type of minimal tool breadth but high depth (ultimately towards just an agent using a computer) you can remove a lot of the scaffolding.
We are almost at state of being able to make the hedgefund name generator for the dozens of places doing same thing. Combinations of flow and whatever suffix {ly, drop, wise...} and boom its probably one of these wrappers.
| Kinda surprised they didn’t run into model collapse problems
Not sure why you would expect this, all the models started doing this as its much more cost effective to get data for post training don't you remember the first grok release where many times it started replies "as a model trained by openai..."
reply