HN user

spunker540

1,208 karma
Posts11
Comments341
View on HN

Could you explain what you mean by this? I thought all of their insane profitability and returns are from crazy margins on their GPUs. I know they started/partnered/invested in some data center businesses, but I thought they were fledgling

The gpu inference happens on anthropics side, but the output of the gpu inference is often many local tool calls. The agent is locally reading files, writing files, compiling programs, downloading artifacts, running tests, invoking CLIs, monitoring logs, inspecting web sites, grepping etc, and can be doing all of that at once, in parallel via subagents, much faster than a single human ever could.

Waymo Premier 1 month ago

now imagine a bus, but its smaller and private only for you. for me that is true freedom! not only can you hop in as drunk as you want, and fall asleep, but you can also control the climate and the music and spread out. and when you arrive the waymo just drives away and you don’t have to think about it ever again in your life

I asked claude to crawl the website and summarize its findings, took about 10minutes. I'm not sure I would've done it faster, but i have no doubt you couldve done it in 5, and grokked the pages faster than an llm too. but anyway heres what claude said:

  Based on what I already saw across those 2,924 pages, here's the summary:

  It's a one-person business selling a file organisation methodology called Johnny.Decimal. Three paid products (personal, business, university/course tier). A substantial blog — 200+ posts, updated weekly. Full documentation for the system. A support knowledge base.

  The technical ambition is higher than the aesthetic suggests. One person built auth, payments, entitlement-gated downloads, a CLI, an API, AI tooling, self-hosted analytics, self-hosted email (Listmonk on PikaPods), personalized search, and keyboard navigation with server-synced state. Then wrote 200 blog posts about using the system in real life. 

  The "Written by humans" footer is not a boast about the font. It's a position statement from someone who has thought carefully about AI, published an essay about it, and is making a deliberate choice. Every word on the site was written by the creator. Whether you agree with the choice or not, that's not the same as someone who slapped a SSG together.

I've hit my claude quota and felt a scary helplessness - but honestly, what if someone took away the toy that is the internet? or the toy that is npm or the toy that is AWS, or the toy that is C# to name a few other toys? plenty of developers can spend their entire careers focused on a single toy and be helpless without said toy.

Whats lost on this thread is these caches are in very tight supply - they are literally on the GPUs running inference. the GPUs must load all the tokens in the conversation (expensive) and then continuing the conversation can leverage the GPU cache to avoid re-loading the full context up to that point. but obviously GPUs are in super tight supply, so if a thread has been dead for a while, they need to re-use the GPU for other customers.

Does SD1.5 suffer from resolution / coherence / complexity issues?

I understand most outputs could be fine tuned for most domains, but still felt sd1.5 had a resolution ceiling, and a complexity ceiling no matter how good the fine tuning

They may have some idea, but they definitely don’t know for sure— there could very well be innocent people on the boat. I’m not sure why arrests are not an option in these cases. It would be great press to announce “x kilos of cocaine captured”, “6 drug smugglers apprehended”

Instead it’s just “boat bombed, terrorists killed, drugs destroyed” with no proof that they’re terrorists or that there are drugs.

I am the same way— I find note taking distracts me such that I end up missing important things, and my notes are usually worthless too. I did pretty well at school with my no-notes approach. Obviously it’s useful to write down assignments, or the occasional verbatim example from the whiteboard. But for me, notes usually hinder more than they help.

I like this idea. While it’s great to have all the music at my fingertips via Alexa + Apple Music (or Spotify etc), it’s actually not very conducive to browsing or recalling all the music and albums I like.

Something physical to browse like this is a pretty fun way to marry the physical world with digital music catalogs.

Not op, but to me this resonates because none of it is “mine”, none of it exists in the real world. There’s a huge difference between the music I physically collected (from libraries, friends, Best Buy, Christmas gifts, used cd stores) and uploaded into my iPod and lived with for years vs music I searched on a whim, listened to for a month while it was in my “recents” and then eventually forgot about once it was pushed out by something else.

I think they are “convincing you” because you are on the opposite side of an opinion.

You’ll notice in this thread, many anti-tracking people are trying to convince pro tracking people, and vise versa. This is common when there are two opposing sides of an argument.

A lot of naysayers here, and I get it’s more about the tactics and messaging more than anything else— but in defense of Duolingo (I’m not a user):

There is a good possibility that in 5 years people have moved on from Duolingo because new ai tutoring apps that can perfectly tailor content to your level and offer unlimited speaking practice sessions, may surpass Duolingo’s “old-fashioned” offering—-the same way Duolingo jumped ahead of websites, which jumped ahead of CDs and cassette tapes, which jumped ahead of books.