spunker540

Thanks for chiming in. Is the issue that they are vibecoded, or specific issues with the changes themselves?

That is the norm where I work , heck I’d say 99% seems low

Do you use coding agents? Just curious bc from my experience using coding agents, open ai’s codex is neck and neck with anthropic’s claude code if not ahead. I wouldn’t agree that OpenAI hasn’t done anything since ChatGPT since codex is my daily driver for software engineering

Do they use Instagram or WhatsApp?

If an LLM can do a rote task for me, I don’t see why I wouldn’t use it

That sounds more like semantic search and vector db.

RAG is simply fetching external data (retrieval) and adding it to LLM context (augmenting) prior to generating a final response.

Any time LLMs do a grep or a web search to answer the query, it’s RAG. Many people use vector db for their own RAG implementation bc of the semantic search benefits.

I agree- I also dont think OS features of the kind that get shipped in these updates are convincing anyone to switch from pc to mac or upgrade their mac sooner. Tons of stuff just changing for the sake of change. Meanwhile LLMs increasingly only need the command line to do everything.

I think the truth is most printed copies of books are rarely read even once, so the idea of preserving them is noble but realistically not that important. And may be outweighed by benefits accrued to the person currently owning and reading it making the most of it and marking to their hearts content.

I admit I don’t mark up my books but I feel tempted to try

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.

Can't a democracy vote and agree to kill innocent people abroad? Not sure those things are discordant.

Last I checked they still do exactly that. Not sure why that case is used as an example when literally every OS bundles a preferred browser

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.

This is a great analogy. Like driving on the freeway, agents are super time efficient, generally safe, but the stakes are high in terms of the worse possible outcomes.

I’m just a swe, but I kinda thought cyber is a good place to be, since the proliferation of insecure vibecoded apps.

I think in the case of supply-constrained GPUs, you can get the opposite of a volume discount. Google has the most capacity of anyone, the fact they’re paying so much per month to spacex is pretty remarkable

I think it depends. Plenty of salaried workers are truly only on the clock when on-site from 9-5

I appreciate the mention of relevant and recent regulation proposals that may affect how future events like this are handled. I understand it did not cause this event but appreciate nonetheless.

I dont know if trade sanctions counts as "sabotage" but point taken.

Saying “ban social media” is a lot like saying to solve lung cancer we must “ban cigarette lighters” when lighters are actually quite useful outside of smoking cigarettes and banning lighters doesn’t really fix the problem.

What exactly would you like banned and how would you define what should be banned and what shouldn’t?

I assume you want FB and Insta banned. What about Reddit? YouTube? Hacker news? Discord? X? Dating apps? Snapchat? WhatsApp? iMessage? Gmail? Just curious where exactly you draw the line, and how you’d implement the ban.

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.

may i ask what number theory breakthrough you refer to? i suspect computing in general or perhaps something more specific?

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.

Yeah there are almost certainly times when it is gen ai and you just didn’t notice.

No. Running a red light is when your possession crosses an intersection while the light is red, and should apply to the registered owner.

Does Google have a better track record when it comes to arbitrarily locking people out of their digital lives?

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.