I just looked up some numbers for UCLA as an example. <45k students (undergrad and grad) and >5k faculty (and another 30k on staff)— so thats a pessimistic ratio of 1 to 9.

If you imagine students take 4 classes per semester and faculty teach 4 per semester… it seems stunningly feasible.

This is a very good argument. We would also learn what features of an App Store add marketable value, and what features are trivial. I imagine the front end isn't very important, but some kind of build certification/verification is. That requires branding, infrastructure and labor. Maybe its easier than I imagine to verify that apps aren't lying about what they do, but as far as I can tell that could well account for some 5% at cost.

On the other hand you trust your bank, for example, so you follow the link on their website and install the App, and the trust came from their own brand.

And people like you will continue to have access to this curated experience. But developers who decide that access to you is not worth the platform fees will be able to pursue an alternative. Why wouldn't your bank remain on the App Store? Does the app store really lack that many dark patterns? (Billions spent gambling every year) Again: There is nothing to stop the walled gardens from being build, but they should be built within a competitive market!

Thanks for the context.

The meat of their argument, in my opinion, relies on this:

> Nevertheless, it is noteworthy that the balance of empirical evidence in this area does not support the claim that social media use has a large impact on users’ political attitudes and behavior.

One such study they provide as a strong representative of the empirical evidence was by researchers collaborating with Meta where they did an RCT to test whether reverse-chronological vs ml algorithm feeds resulted int different political beliefs. I haven't looked into the study, but on its face that's an insanely stupid design; if my youtube feed became only reverse chronological one day I'd just open a different app after 40 seconds.

Content platforms' main product is their behavior-modulating feeds, their ability to hold your eyes for 4 hours a day. The idea that this wouldn't be effecting our politics is insane.

You totally missed the point.

Critics: CAHSR is a bad idea and should be killed.

Author: Here's the story of how hard it is to build this novel type of large scale infrastructure, especially when people don't strongly support it.

The dysfunction is real, but the actual point was, "How much of this dysfunction and how much was just what you get with half-assed support?"

You're interpretation of that "no track has been laid" quote betrays that you aren't evaluating this fairly. You really don't understand that laying rocks and steel on a path is easier than building whole bridges and viaducts.

How does 'stress of being in crowds' in big cities indicate all those people would prefer to drive? Is stuck in traffic not the 'stress of being in crowds'?

The stress of 2 ton machines flying around at 120 km/hr and operated by the angry and impatient-- that is simply not worth it. You think automated trains with 2 minute headways in a network covering 85% of local journeys at $3/ride would be worse than a Taxi at $1/minute with traffic?!!

A lot of being "car brained" is not realizing that even decent transit by global standards would be far far better than the subsidized freeway only-if-you-can-drive $8,000/year horrorshow of the US/Canada. No one's going to take away your fun cars, but a system maximizing freedom needs to account for the young, old, disabled, drunk, poor, and motivated to read instead of drive. Mass transit is freedom. Cars are consumption-politics/corporate-power.

There is definitely an ethic associated with 'being informed'; I remember being told to read the news as a kid and it felt like vegetables.

Scroll media is fast food, and fine dining is books or long form sub-stack-- which cost more money but also will-power. The question of how scroll media can deliver high quality information is similar to asking how drive through can serve vegetables. I think it comes down to the fact that you can't cultivate taste unless people are paying with will-power.

Yes, surely a sarcastic reductio ad absurdum of what was was said will inspire dialogue. I think the GP's point is that their investing in new distribution channels could mean ROI in models has diminished significantly. Incidentally, I disagree with GP that's what this means-- this is another investment in brand awareness, AND data for multi-modal/audio. They might have gotten to 1080p for text chat but definitely not for voice chat.

Nothing more absurd than your response. OpenAI has a large engineering staff, it’s foolish to say they are all working on advancing models. The folks working on ChatGPT are going to continue working on ChatGPT. Let’s not even forget that O1 just got released recently.

Nothing I said was absurd in response to making an unsupported idea that model development has plateaued.

Well that's terrible news. Currently building a product for the same market (completely different solution if the nephew comment is correct about your product). I'm already not thrilled to be in ed-tech selling to instructors with admin's money. I thought at least the instructors would have some enthusiasm to solve the problem. Shit.

Great site. Small feedback: There's a category 'Closure'-- I'm not sure if that's something I don't know about, but it definitely isn't for jobs using 'Clojure'.

Oh wow that’s not good, thank you so much for pointing that out!