frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Open in hackernews

Altman fluffs superintelligence to save humanity as OpenAI slashes prices

https://www.theregister.com/2025/06/11/openais_sam_altman_superintelligence/
34•rntn•1d ago

Comments

ulfw•1d ago
I am having a hard time figuring out when Elon and when Sam said something nowadays
notrealyme123•1d ago
This has to be the most diplomatic way to write "This guy sounds like he's on drugs".
i_am_proteus•1d ago
>GPT-4o query consumes around 0.3 watt-hours for a typical text-based question, though this increases substantially to 2.5 to 40 watt-hours for queries with very long inputs.

Given a context window of 128 kilotokens, and how the energy usage range (0.3 to 40 Wh) increases by a factor of 133 from "single query" to "worst case," this suggests linear scaling of energy usage with context length.

Given the common user pattern of "chat with the LLM" the 0.3 Wh figure seems reductive. As the chat grows in length, the length of the "query" becomes the length of the entire chat.

Assuming context compression for long chats, the figure we're looking for per-question energy use should be the energy use for a query half the length of max context tokens.

For GPT-4o: 20 Wh

enqk•1d ago
If you only have two points you would always assume linear. But what if it’s quadratic, like this article claims?

https://www.timdavis.com/blog/scale-or-surrender-when-watts-...

i_am_proteus•23h ago
Good point!

The good news would be that GPT-4o average energy usage per query would be lower than 20 Wh.

The bad news is that there's a quadratic increase in energy usage with the increase in a model's maximum context window. GPT 3.5 -> GPT 4 was an increase from thousands of tokens to hundreds of thousands of tokens.

andrewstuart•1d ago
I’m interested in how good it is at coding, only a little interested in other stuff.
myko•23h ago
I'm interested in when it can do my laundry and put away my dishes. Maybe even move furniture / organize my stuff.
berbec•22h ago
I don't want AI to code while I'm scrubbing the floors. I want AI to scrub the floors while I code.
mrweasel•1d ago
I'm sorry, but I'm not buying into that level of optimization. Earlier this year they struggled to make a profit on Pro at $200, granted that was for unlimited usage, but now they've optimized the design/implementation sufficiently to slash prices by 80%?
helloplanets•23h ago
I absolutely believe inference is cheap and efficient, but the net impact of training the model should be also added to the equation. Not just the cost of inference.

Is this information readily available somewhere?

rvz•23h ago
> "Maybe we will go from solving high-energy physics one year to beginning space colonization the next year; or from a major materials science breakthrough one year to true high-bandwidth brain-computer interfaces the next year," he said. Maybe. Or maybe not. We note that fellow futurist Elon Musk, who once cautioned about releasing the AI demon, predicted in 2016 that humans would land on Mars by 2025.

"Maybe"

What about the definition of "AGI" being changed again to mean something else? What does "AGI" actually mean anymore or was it hijacked again for the purpose of manipulation?

But this is the most important sentence of the entire article:

> Tech leaders simply pay no cost for misprediction.

They don't care if they are wrong on their predictions. So focus on what they are doing rather than what they are saying.

chneu•20h ago
Business man lies to sell product.

Jemalloc Postmortem

https://jasone.github.io/2025/06/12/jemalloc-postmortem/
500•jasone•10h ago•131 comments

If the moon were only 1 pixel: A tediously accurate solar system model (2014)

https://joshworth.com/dev/pixelspace/pixelspace_solarsystem.html
77•sdoering•3h ago•10 comments

Frequent reauth doesn't make you more secure

https://tailscale.com/blog/frequent-reath-security
904•ingve•16h ago•397 comments

Rendering Crispy Text on the GPU

https://osor.io/text
240•ibobev•9h ago•76 comments

A receipt printer cured my procrastination

https://www.laurieherault.com/articles/a-thermal-receipt-printer-cured-my-procrastination
946•laurieherault•1d ago•498 comments

A Dark Adtech Empire Fed by Fake CAPTCHAs

https://krebsonsecurity.com/2025/06/inside-a-dark-adtech-empire-fed-by-fake-captchas/
168•todsacerdoti•13h ago•65 comments

iPhone 11 emulation done in QEMU

https://github.com/ChefKissInc/QEMUAppleSilicon
309•71bw•20h ago•28 comments

Slow and steady, this poem will win your heart

https://www.nytimes.com/interactive/2025/06/12/books/kay-ryan-turtle-poem.html
37•mrholme•6h ago•21 comments

Show HN: Tritium – The Legal IDE in Rust

https://tritium.legal/preview
212•piker•23h ago•120 comments

Three Algorithms for YSH Syntax Highlighting

https://github.com/oils-for-unix/oils.vim/blob/main/doc/algorithms.md
31•todsacerdoti•9h ago•9 comments

Maximizing Battery Storage Profits via High-Frequency Intraday Trading

https://arxiv.org/abs/2504.06932
248•doener•1d ago•237 comments

Rust compiler performance

https://kobzol.github.io/rust/rustc/2025/06/09/why-doesnt-rust-care-more-about-compiler-performance.html
220•mellosouls•3d ago•183 comments

Show HN: Tool-Assisted Speedrunning the Boring Parts of Animal Crossing (GCN)

https://github.com/hunterirving/pico-crossing
118•hunterirving•22h ago•16 comments

Why does my ripped CD have messed up track names? And why is one track missing?

https://www.akpain.net/blog/inside-a-cd/
135•surprisetalk•20h ago•120 comments

Show HN: McWig – A modal, Vim-like text editor written in Go

https://github.com/firstrow/mcwig
123•andrew_bbb•22h ago•13 comments

Urban Design and Adaptive Reuse in North Korea, Japan, and Singapore

https://www.governance.fyi/p/adaptive-reuse-across-asia-singapores
35•daveland•9h ago•5 comments

Chatterbox TTS

https://github.com/resemble-ai/chatterbox
621•pinter69•1d ago•182 comments

Microsoft Office migration from Source Depot to Git

https://danielsada.tech/blog/carreer-part-7-how-office-moved-to-git-and-i-loved-devex/
333•dshacker•1d ago•257 comments

Worldwide power grid with glass insulated HVDC cables

https://omattos.com/2025/06/12/glass-hvdc-cables.html
93•londons_explore•15h ago•62 comments

Solving LinkedIn Queens with SMT

https://buttondown.com/hillelwayne/archive/solving-linkedin-queens-with-smt/
114•azhenley•19h ago•36 comments

The curse of Toumaï: an ancient skull and a bitter feud over humanity's origins

https://www.theguardian.com/science/2025/may/27/the-curse-of-toumai-ancient-skull-disputed-femur-feud-humanity-origins
57•benbreen•13h ago•18 comments

Dancing brainwaves: How sound reshapes your brain networks in real time

https://www.sciencedaily.com/releases/2025/06/250602155001.htm
166•lentoutcry•4d ago•60 comments

They Asked an A.I. Chatbot Questions. The Answers Sent Them Spiraling

https://www.nytimes.com/2025/06/13/technology/chatgpt-ai-chatbots-conspiracies.html
4•cainxinth•58m ago•1 comments

First thoughts on o3 pro

https://www.latent.space/p/o3-pro
163•aratahikaru5•2d ago•146 comments

Quantum Computation Lecture Notes (2022)

https://math.mit.edu/~shor/435-LN/
144•ibobev•4d ago•46 comments

Show HN: I wrote a BitTorrent Client from scratch

https://github.com/piyushgupta53/go-torrent-client
118•piyushgupta53•6h ago•35 comments

Show HN: GetHooky – a language-agnostic Git hook manager

https://ezpieco.github.io/GetHooky/
11•Ezpie•8h ago•3 comments

Helion: A modern fast paced Doom FPS engine in C#

https://github.com/Helion-Engine/Helion
161•klaussilveira•2d ago•61 comments

US-backed Israeli company's spyware used to target European journalists

https://apnews.com/article/spyware-italy-paragon-meloni-pegasus-f36dd32106f44398ee24001317ccf2bb
645•01-_-•19h ago•320 comments

Roundtable (YC S23) Is Hiring a President / CRO

https://www.ycombinator.com/companies/roundtable/jobs/wmPTI9F-president-cro-founding
1•timshell•14h ago