Asmit

AGI will be the ultimate monument to human pride, and simultaneously its ultimate undoing

Incredibly excited to release GPT-5.1 Pro to the world! It’s great at tackling the hardest, messiest problems with clearer, more comprehensive responses. I’m eager to see you throw your toughest problems at it. Please try it and share what you think :)

@sporadica still got a grazer in me

matcha just tastes like dirt why do y’all people like this stuff

@rcx86 could have dropped the "unit", to make it more dense.

Information density per unit sentence is a very high indicator of intelligence

@rohanpaul_ai It gets dicey when the sources cited by the likes of perplexity/AI mode/searchGPT are AI generated themselves. As long as it's only used for paraphrasing/ readability it should be fine.

The paper audits US newspapers and finds AI writing common, uneven, and mostly undisclosed. 😃 9% of new articles show AI use. The researchers looked at 186K articles from 1,528 newspapers across the US. At the New York Times, Washington Post, and Wall Street Journal, opinion pieces use AI about 6.4x more than news. Almost no one admits it though. Only 5 out of 100 AI-written stories actually said they used AI. 😀 ---- Paper – arxiv. org/abs/2510.18774 Paper Title: "AI use in American newspapers is widespread, uneven, and rarely disclosed"

@raphaelschaad You can ask Siri iPhone questions.

Safari now forcing me to do three taps to close one tab surprisingly is one of my biggest UX annoyances of iOS 26. iOS 18 vs. 26:

@yacineMTB Is ur username really कछे 🩲

leaving behind millions in unvested stock is a canon event that every staff engineer has to go through. people don't know how hard it is to be us

@tszzl This is why humans have conquered the seas, crossed the lands. Couldn't just sit and enjoy some bread.

friend has pivoted to a chat first website. the pendant is now an "upgrade" available on a secondary page that used to be their home page. this is great change imo, getting the core of the product to more users, with the hardware being a secondary purchase. kudos @AviSchiffmann

@khoomeik @tszzl It would check out tho. When there are no adventures, no wishes, nothing to conquer for the sapiens. A perfectly created AI slop machine will quench the brain's thirst to be unsettled. Stimulations to move our brains around, but not move us anywhere.

@tszzl they’re definitely mutually compatible, with carbons wireheading themselves and silicons engineering themselves to kardashevmaxx this isn’t a particularly ideal scenario tho no? i guess it’d also suck if we use the dyson sphere to power sloptok v527.63.11

which way, anon?

@sporadica Even after having this prior, there is an uncontrollable primal urge to yeet patience and have a fleeting shot at reaching the exit cave first. Definitely some evolutionary slag.

the plane just landed, you are in the last row, and you rushed to stand up as soon as possible now you must stand for 20 minutes to get off of the plane 2 seconds quicker meanwhile i, i take pleasure in your suffering

@tszzl Our fractal minds have a tendency to see self similar orders. The similarity in the abstraction of a man, and a state is an exhibit.

“By art is created that great Leviathan, called a Commonwealth or State—(in Latin, Civitas) which is but an artificial man.”

@_nonfigurativ_ @sanity_io Its giving Pantheon

mem_brane: Inside an artificial mind (installation simulation)

A few of us at Georgia Tech are building Vajra, an open‑source inference engine built for the next wave of serving problems: * Multimodal streaming first – Low‑latency realtime video, audio pipelines * Highly distributed computing - Managing 1000s of GPUs * Hierarchical Graph Scheduling - A new computation model for large scale distributed computing We just published the vision  [project-vajra.github.io/blog/vajra-vis…](project-vajra.github.io/blog/vajra-vis…) Who we’re looking for We need mission driven systems hackers who can own subsystems and want to push the frontier of computing. If that sounds like your idea of fun, reach out to me at ameyagrawal@gatech.edu with subject "Vajra".

The bitter lesson of AI infra: The hardest part about building faster LLM inference systems is not designing the systems, but rather it is evaluating if the system is actually faster! 🤔 This graph from a recent top systems venue paper about long-context serving shows average normalized input token latency for a trace with both short and 100K+ token requests. System X looks like a clear win: lower normalized latency and higher request rates. But normalized metrics can obscure the actual user experience: at those rates, long inputs see >2hr delays to the first token! Let’s do the math!🧮

Super long-context models with context window spanning millions of tokens are becoming commonplace (@GoogleDeepMind Gemini, @xai Grok 3, @Alibaba_Qwen Qwen2.5). But efficiently serving these models is tough, especially alongside short requests. Head-of-Line (HOL) blocking becomes a major issue, hurting latency for everyone. We present Medha, a system designed to handle this mix efficiently. Achieving 30x lower latency, and 5x higher throughput compared to the state-of-the-art. Full paper: arxiv.org/pdf/2409.17264. 🧵

This Apple Intelligence feature hits the spot for inquisitive but impatient types like me. Information snacking.

@emilymbender @AravSrinivas And as far as, incorrect summarisation is concerned, most blogs and pages are second/third order pieces of information which themselves are an attempted summarisation of the facts. LLMs are superior in playing with words as of now.

@emilymbender Most of the “LLM-driven” searchbots are RAG based, aligned to only emit lines that can be cited from external sources. I think @AravSrinivas mentioned that you can imagine this search response to be like writing an academic paper, where each line refers to a citation or itself.

As OpenAI and Meta introduce LLM-driven searchbots, I'd like to once again remind people that neither LLMs nor chatbots are good technology for information access. A thread, with links: >>