Simon Econ @SimonEconTips

DeepSeek has launched DeepSeek-V3.2-Exp, an experimental upgrade to V3.1-Terminus - introducing DeepSeek Sparse Attention (DSA) for faster, more efficient long-context processing with minimal quality loss. -The model matches V3.1 in performance while cutting API prices by over…

keyboard_arrow_left Previous keyboard_arrow_right Next

DeepSeek @deepseek_ai

13 hours ago

DeepSeek has launched DeepSeek-V3.2-Exp, an experimental upgrade to V3.1-Terminus - introducing DeepSeek Sparse Attention (DSA) for faster, more efficient long-context processing with minimal quality loss. -The model matches V3.1 in performance while cutting API prices by over… https://t.co/jdGUamsyQI

179 710 5K 759K 793

What xAI does is absolutely insane…it’s a total game reset Grok 4 Fast delivers GPT-5 (High) performance - for just 5% of the cost (95.7% cheaper)... Insane value If you're still no using Grok 4 Fast, you're just literally burning money Grok 4 Fast costs - Input tokens:…

The jump from "agents are nowhere close to working" to "okay, narrow agents for research and coding work pretty well" to (very recently) "general purpose agents are actually useful for a range of tasks" has been quick enough (less than a year) so that most people have missed it.

Since we’re -3pp away from human level parity in 44 industries but haven’t seen mass job displacement yet, what can we deduce? * AI is augmenting humans, not replacing? * AI is great, but last mile product integration isn’t there yet? (App layer) * Expert level benchmarks can…

OpenAI @OpenAI

4 days ago

Since we’re -3pp away from human level parity in 44 industries but haven’t seen mass job displacement yet, what can we deduce? * AI is augmenting humans, not replacing? * AI is great, but last mile product integration isn’t there yet? (App layer) * Expert level benchmarks can…

14 53 627 313K 118

Download Image

OpenAI have 9x'd their capacity this year, and have a goal to 125x further by 2033, thus exceeding current energy capacity of the whole of India - as per @alexeheath reporting of @sama's internal slack note. One thing to note is that 125x energy capacity increase understates the…

maybe we'll reach a point where we can brute-force superintelligent AI into existence. even without new innovations, scaling compute could keep pushing intelligence higher it might not be true AI, but with enough compute, systems could simulate it for most robotic tasks

The Singularity Is In Our Destiny AI was inevitable! It’s more of a discovery than an invention The patterns of the universe have always been there waiting for someone to understand them Digital intelligence simply encodes and generates these patterns with probabilities…

Accenture plans on 'exiting' staff who can't be reskilled on AI amid restructuring strategy cnb.cx/4nIUEaw @CNBC @CNBCMakeIt

Paving the way for unlimited context windows. Introducing @zml_ai /attnd: sparse logarithmic attention, on CPU, over UDP, faster than GPU.

$SAP CFO: "I am sometimes baffled when I hear from some investors that all customers will write their own software themselves. We're doing the same. So now the question is, who has the bigger relative advantage? Will it be a not-so-proficient software department of an industrial…

Eric Schmidt says data centers need about 92 gigawatt more capacity, with almost no new nuclear projects.

BP Plc said that oil demand is going to keep growing for the rest of this decade, rowing back on its prior projection that the high point could come as soon as this year. #oott bloomberg.com/news/articles/…

OpenAI wants to create a "factory" that will eventually be able to create 1 GW of new AI infrastructure per week. For reference, yesterday's NVIDIA deal was for "just" 10 GW total. OpenAI has been "hard at work on this" and "believes it is possible". Key details regarding…

Sam Altman @sama

6 days ago

OpenAI wants to create a "factory" that will eventually be able to create 1 GW of new AI infrastructure per week. For reference, yesterday's NVIDIA deal was for "just" 10 GW total. OpenAI has been "hard at work on this" and "believes it is possible". Key details regarding…

677 707 5K 1.5M 2K

This is the 4th shot! A small upgrade on Qwen3-Coder, but should be a good improvement in user experience. Its performance on Terminal Bench is better either with Qwen Code or Claude Code. Its coding agent capabilities are enhanced, which achieves 69.6 in SWE-Bench-Verified, and…

Qwen @Alibaba_Qwen

6 days ago

This is the 4th shot! A small upgrade on Qwen3-Coder, but should be a good improvement in user experience. Its performance on Terminal Bench is better either with Qwen Code or Claude Code. Its coding agent capabilities are enhanced, which achieves 69.6 in SWE-Bench-Verified, and…

48 148 1K 154K 232

Download Image

Sonnet 4 Now Has Serious Competition GPT-5 Codex just launched in API. 🚀🚀 Although it may not seem like it, this is a watershed moment in the history of AI. Claude has been so dominant in the agentic space that it has been impossible to displace! Codex will present serious…

When I say Swiss-Army knife, I mean it. Here's another video showing how to automatically keep yourself updated with github issues on a repo of interest. Watch it go through the issues and do the magic! The folks at @ChromiumDev have suddenly up-leveled everyone's game.

Delip Rao e/σ @deliprao

6 days ago

When I say Swiss-Army knife, I mean it. Here's another video showing how to automatically keep yourself updated with github issues on a repo of interest. Watch it go through the issues and do the magic! The folks at @ChromiumDev have suddenly up-leveled everyone's game. https://t.co/zD58pVZANf

27 134 1K 231K 1K

Download Video

🚀 Introducing Qwen3-Omni — the first natively end-to-end omni-modal AI unifying text, image, audio & video in one model — no modality trade-offs! 🏆 SOTA on 22/36 audio & AV benchmarks 🌍 119L text / 19L speech in / 10L speech out ⚡ 211ms latency | 🎧 30-min audio…

Oracle’s $ORCL Larry Ellison: “Training AI models is a gigantic multitrillion-dollar market. It's hard to conceive of a technology market as large as that one. But if you look close, you can find one that's even larger. And it's the market for AI inferencing.” $NVDA $AMD

sam altman new blog on: Abundant Intelligence - maybe with 10 GW of compute, AI can figure out how to cure cancer - openAI wants to create a "factory" that will eventually be able to produce 1 GW of new AI infrastructure per week this is the Apollo program of our generation,…

Sam Altman @sama

6 days ago

sam altman new blog on: Abundant Intelligence - maybe with 10 GW of compute, AI can figure out how to cure cancer - openAI wants to create a "factory" that will eventually be able to produce 1 GW of new AI infrastructure per week this is the Apollo program of our generation,… https://t.co/kP3jUJDmkc

677 707 5K 1.5M 2K

Sam Altman’s master plan to win the AI race is to corner all the funding, starving the competition to death.