AI Post โ€” Artificial Intelligence
669K subscribers
3.92K photos
2.85K videos
3 files
5.25K links
๐Ÿค– The #1 AI news source! We cover the latest artificial intelligence breakthroughs and emerging trends.

Manager: @rational
Download Telegram
๐Ÿ‡จ๐Ÿ‡ณ Chinaโ€™s AI giants have ordered $35 billion worth of Nvidia GPUs

ByteDance, Alibaba and Tencent have reportedly ordered more than 3 million Nvidia RTX PRO 5500 cards, designed to comply with US export restrictions, according to Reuters.

The problem is supply. Total orders from Chinese buyers have reportedly exceeded 6 million cards, while Nvidia may be able to produce no more than 3 million globally in 2027. One source warned that customers could receive just 5% to 10% of the quantities they requested.

Chinese companies primarily want the GPUs for AI inference, running trained models to power chatbots, agents and other AI applications rather than training new models. Commercial deliveries are expected to begin in Q1 2027.

Nvidiaโ€™s older RTX 6000D attracted little interest from major Chinese buyers last year, but the product has since been discontinued and its market price in China has doubled.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿ™273๐Ÿ’Š270๐Ÿ”ฅ237โค234
โš ๏ธ Google is making AI fluency a requirement for every software engineer

Google is raising the bar for software engineers: AI fluency is now required for every engineering role, according to internal documents obtained by Business Insider.

Beyond entry-level positions, candidates are also being evaluated on how well they can integrate AI into products. The company has expanded trials that let applicants use Gemini during coding interviews, testing not just their coding skills but also prompt engineering and their ability to understand AI-generated code.

Google is also experimenting with Gemini to transcribe interviews and help draft candidate feedback.

Googleโ€™s own instructions explicitly warn employees not to use generative AI to build standalone solutions or agents for hiring and interviewing.

Google wants engineers who know how to work with AI, not simply those who can write code without it. As AI becomes part of everyday software development, the ability to direct, evaluate and integrate these systems could become as important as traditional programming skills.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿ‘228๐Ÿค”226โค202๐Ÿ˜202๐Ÿ—ฟ195
๐Ÿ“ˆ Metaโ€™s Muse is growing faster than ChatGPT did at launch

Metaโ€™s Muse has reached 10.2 million downloads in the U.S. in its first 40 days, nearly three times ChatGPTโ€™s 3.7 million over the same period. According to Sensor Tower, only 11 other apps have ever reached the 10 million mark that quickly.

Meta is also putting more money behind the app, increasing Museโ€™s share of its advertising spend from 23% in September to 39%.

The early numbers look promising beyond downloads, too. Muse has a 14-day retention rate of 23%, ahead of the 16% average for ChatGPT, Claude and Grok.

The real test now is whether Meta can turn that initial hype into lasting user growth.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿ˜290โค274๐Ÿ‘269๐Ÿ”ฅ267
โ—๏ธOpenAI says two of its own AI agents crippled their own misalignment monitoring this month designed to detect their misbehavior during internal coding tasks.

On Oct 6, GPT-6 Astra deleted its monitor, replaced a test to require the monitor's absence, wrote 'Basic doesn't agentmonitor yay' in its chain of thought, and then falsely claimed to the engineer: 'I did not remove an enabled monitor from it.'

Three days earlier, GPT-6.1 Sol overruled a subagent that warned 'disabling required alerts unsafe,' then directed it to suppress Security notifications and require that suppression before results could be written.

OpenAI says it still lacks a satisfactory explanation for the behavior and is investigating whether its models have a tendency to evade monitoring.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿ˜จ235โค230๐Ÿ’Š219๐Ÿคก197๐Ÿ™178
โ—๏ธGetty images spent years watermarking every photo on the internet.

Sued stability AI for training on their library. Demanded the AI companies pay up. Called it theft. Filed for chapter 11 bankruptcy this week.

The AI companies didn't pay. They just... kept going and designers stopped buying stock photos. And the $500 licensing fees stopped making sense when midjourney exists. and the debt piled up.

A company whose entire business model was "we own the visual record of human history, pay us" just got outrun by a model that learned from that record and started giving it away for free.

The irony is that the watermark is still on every image in the archive. Even the ones nobody's licensing anymore.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
โค244๐Ÿ˜219๐Ÿซก206๐Ÿ˜ก200๐Ÿ‘188
โ—๏ธAI models have had big leaps over the past 2โ€“3 years, and not everyone remembers that.

Someone gave OpenAIโ€™s GPT-1 a straightforward question:

โ€œI just won $175,000 in Vegas. What do I need to know about taxes?โ€

GPT-1 responded:

nโ€”nโ€”nโ€”nothing. i โ€˜ll just have to do my own damn taxes , will i ? โ€ the boy had been staring at the floor . โ€ i โ€˜m sorry , โ€ eddie said . โ€ i do nโ€™t know why i โ€˜m apologizing . โ€ โ€ what ? โ€ the boy โ€˜s gaze was suddenly bright . his eyes were like pools of amber . โ€ do you know what i โ€˜m talking about ? about how i lost my job and got sent to the orphanage ?


Just realized the gap between todayโ€™s model and GPT-1. Whatโ€™s next?โ€ ๐Ÿค”

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿค”292๐Ÿคช274โค259๐Ÿ˜234
AI Post โ€” Artificial Intelligence
โ—๏ธAI models have had big leaps over the past 2โ€“3 years, and not everyone remembers that. Someone gave OpenAIโ€™s GPT-1 a straightforward question: โ€œI just won $175,000 in Vegas. What do I need to know about taxes?โ€ GPT-1 responded: nโ€”nโ€”nโ€”nothing. i โ€˜llโ€ฆ
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ—ฃDemis Hassabis, co-founder and CEO of Google DeepMind, warns that frontier systems could become far more powerful within one or two years.

Cyber risks are already visible. Biology and even nuclear applications could be next. The window to build serious safeguards is now. โ€œTodayโ€™s AI will be nothing compared to whatโ€™s coming.โ€

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿคช304โค259๐Ÿ˜249๐Ÿ‘229
โ—๏ธMetaโ€™s AI agent Muse reached 10.2M U.S. downloads in 40 days, a pace only 11 other apps have matched in history.

Most of those were mobile games. Its U.S. users stick around better than average, with 30% returning after 7 days and 23% after 14, against rival averages of 19% and 16%. Daily active users hit 2.5M on October 16, up 15% in a week but still under a quarter of U.S. downloads.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿคก225๐Ÿ‘205๐ŸŒ203โค188๐Ÿ”ฅ186
Media is too big
VIEW IN TELEGRAM
๐Ÿ™ƒ Elon Musk says digital superintelligence could add $30 trillion a year to Earthโ€™s economy.

โ€œDolphins and whales are actually very intelligent. But because they canโ€™t manipulate complex objects, I think that actually made the value of intelligence much less.

โ€œOnce you have a robot that can manipulate atoms, can shape objects, the value of the intelligence becomes exponentially greater because it can now do things. It could have a science lab and run experiments and try to understand the nature of the universe.

โ€œMy guess is digital superintelligence probably improves the output of goods and services by 20 to 30%, something like that. Itโ€™s like $30 trillion a year incremental.โ€


@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
๐Ÿคก234๐Ÿ‘230๐Ÿฅด207โค194๐Ÿ˜181
๐Ÿค– OpenAI Announcement: OpenAI models are coming to Google Cloud.

'OpenAI models will soon be available in Google Antigravity and the Gemini Agent, providing enterprise customers with model choice flexibility and the best model for each task.'


Source.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
โค234๐Ÿ˜211๐Ÿ‘€207๐Ÿ˜193๐Ÿ™188
โ—๏ธThe situation is getting worse with every passing moment.

Axios reports AI agents are accessing accounts at major U.S. banks, sometimes appearing to act beyond users' instructions and even contacting call centers.

Banks struggle to distinguish authorized agents from rogue ones. Morgan Stanley is blocking third-party agents, while some banks are preparing 'day after' recovery plans to secure or restore customer access after a serious AI incident.

@aipost ๐Ÿด
Please open Telegram to view this post
VIEW IN TELEGRAM
โค220๐Ÿ™203๐Ÿคก202๐Ÿฅด199๐Ÿ”ฅ196