AI Post — Artificial Intelligence
Open in Telegram
🤖 The #1 AI news source! We cover the latest artificial intelligence breakthroughs and emerging trends. Manager: @rational
Show more2025 year in numbers

979 978
Subscribers
-50524 hours
-3 5877 days
-15 75230 days
Posts Archive
Photo unavailableShow in Telegram
🔔 McKinsey says AI tools could technically automate about 57% of work hours in the U.S., but that most jobs will change instead of disappearing.
A lot of tasks inside jobs can be automated, but full roles usually mix those tasks with human parts like context, trust, and on the spot judgment that AI still struggles with. Job postings already reflect this shift, with demand for “AI fluency” growing about 7x in 2 years
Highly routine brain work like simple accounting flows or boilerplate coding is most exposed, because AI can already handle document prep, pattern matching, and basic research. To unlock the projected 2.9 trillion dollars per year of value by 2030 from AI, companies have to redesign how work is split between humans, agents, and robots instead of just bolting AI onto old processes.
Source.
AI Post 🪙 | Our X 🥇
👍 8❤ 7👾 5
⚠️ China has overtaken the US in the global market for open AI models
A study by MIT and Hugging Face found that Chinese-made open-source AI models have overtaken American developers in global downloads for the first time, capturing 17 percent of the market compared to 15.8 percent for U.S. creators in the past year.
DeepSeek and Alibaba's Qwen dominate Chinese downloads, while U.S. giants like Google, Anthropic and OpenAI remain focused on closed, proprietary systems.
Article. Paper.
AI Post 🪙 | Our X 🥇
👍 18❤ 13👀 6🔥 1
01:51
Video unavailableShow in Telegram
Ilya Sutskever predicts that as AI becomes visibly powerful, human behavior will shift in new ways
Frontier labs will work together on safety, and governments and the public will push to act. Once AI feels truly powerful, the industry's mindset will shift, and companies will become more cautious.
AI Post 🪙 | Our X 🥇
IMG_1587.MP486.71 MB
👾 12👍 10🤔 4😢 4😁 2
Photo unavailableShow in Telegram
⚠️ The US economy is now BUILT on AI:
US private nonresidential fixed real investment in data centers has soared +300% over the last 3 years. During the same period, inflation-adjusted nonresidential investment in structures excluding data centers has barely seen any growth.
These are offices, shopping centers, hotels, warehouses, factories, manufacturing facilities, and other commercial buildings. Meanwhile, AI-related spending has contributed 62.5%, or 1 percentage point, to the 1.6% US GDP growth in the first 6 months of 2025.
Without AI, the US would be in a recession.
AI Post 🪙 | Our X 🥇
🤪 18❤ 7👍 5👾 1
12. "How would I solve this if I only cared about physics, not politics?"
Musk removes social friction from problem solving. “I need to restructure my team. How would I solve this if I only cared about physics?” AI gives the uncompromised version first, then you can soften it.
13. "If I had to achieve this 10 times faster, what would I do?"
Extreme deadlines force extreme creativity. "I'm planning a 6 month launch. If I had to achieve it 10 times faster, what would I do?" AI re-engineers your process into speed.
14. "What would this look like if it had to scale to millions?"
Musk thinks in orders of magnitude. "I'm creating a small online course. What would this look like if it had to scale to millions?". AI finds failure points and opportunities early.
15. "Which part of this solution creates the most leverage?"
Musk always builds the part that changes everything. "My marketing strategy feels scattered. What creates the most leverage?". AI reveals the multiplier.
The secret multiplier: Stack these together
Use this prompt: "Break my problem into fundamental truths, strip all assumptions, find the optimal solution, identify hidden constraints, and rebuild the idea from first principles.". AI becomes a first principles engineer inside your creative process.
AI Post 🪙 | Our X 🥇
👀 11👍 7❤ 6🤪 3🤔 2
8. "What part of this is actually impossible and what part just feels impossible?"
Most limits are emotional. "I'm considering a pivot. What part of this feels impossible instead of being impossible?". AI separates physics from fear.
9. "What is the minimum viable breakthrough?"
Not minimum viable product. Minimum viable breakthrough. "I want to grow a YouTube channel. What is the minimum viable breakthrough?". AI finds the foundational leap that unlocks growth.
10. "If I restarted this entire project today, knowing what I know now, what would I build?"
Clean slate thinking. "I have a messy SaaS that evolved randomly. If I restarted today, what would I build?". AI reveals the elegant version buried inside the bloated one.
11. "What are the hidden constraints I'm not questioning?"
Most problems hide fake walls. "My agency is stuck at 20 clients. What hidden constraints am I not questioning?". AI exposes self imposed ceilings.
Next part: ⤵️⤵️
❤ 10👍 6🤪 3😐 1
Photo unavailableShow in Telegram
Elon Musk's actual thinking framework was used as AI prompts.
It's the closest thing to having a billionaire engineer rip apart your ideas and rebuild them from physics. Here are the 15 prompts that changed how I solve problems:
1. "What are the physics of this problem?"
Musk strips everything to objective reality. "I'm struggling to grow my newsletter. What are the physics of this problem?". AI reveals the hard constraints, the real forces, and the non-negotiable bottlenecks.
2. "If I couldn’t rely on existing assumptions, how would I solve this?"
Assumptions are invisible cages. "My pricing model is based on what competitors do. If I removed all assumptions, how would I solve this?". AI breaks the mental autopilot.
3. "What are the problem’s fundamental components?"
Musk breaks problems into atoms. "My business is plateauing. What are the fundamental components of this problem?". AI decomposes your challenge into parts that can actually be solved.
4. "What would the optimal solution look like if cost didn’t exist?"
Constraints ruin creativity too early. "I need to redesign onboarding. What would the optimal solution look like if cost didn’t exist?". AI generates the ideal version first, then works backward.
5. "If I were forced to cut 90 percent of this, what would remain?"
Brutal prioritization. "My product roadmap has 50 items. If I had to cut 90 percent, what remains?". AI forces clarity through elimination.
6. "If this failed completely, what would be the root cause?"
Musk starts with failure to engineer success. "I'm launching a new product. If it failed, what would be the root cause?". AI predicts the exact failure modes in advance.
7. "What would a solution look like if I ignored industry norms?"
Musk bypasses entire industries by refusing to copy them. "I'm starting a consulting business. What would a solution look like if I ignored industry norms?" AI gives you the non obvious path.
Next part: ⤵️⤵️
🙏 8👍 5🔥 3😐 3😁 1
Photo unavailableShow in Telegram
⚡️ DeepSeek drops Math-V2 and it’s a monster
No hype, no teasers, DeepSeek just pushed out Math-V2 almost two years after the original. No one was waiting for it… and yet it quietly arrived with absurd numbers.
Performance snapshot:
• Gold-medal strength on IMO 2025 and China’s CMO 2024.
• 118/120 on Putnam 2024 , a famously brutal university-level contest.
• Sparse benchmarks (no MATH, GSM8K, AIME), but IMO-ProofBench shows it outscoring GPT-5 and matching Gemini 2.5 Pro on hard proofs. Now everyone’s watching what Gemini 3, Grok 4, and GPT-5.1 will do next.
What’s inside the system:
• Built on DeepSeek-V3.2-Exp-Base.
• Architecture = Generator + Verifier loop.
• Solutions are not produced in one pass. Instead:
1. Generator drafts a proof.
2. Verifier inspects every step, flags errors.
3. Generator rewrites and improves.
4. Repeat, up to 16 refinement rounds.
Massive test-time compute:
• Each iteration evaluates up to 64 parallel hypotheses.
• A single problem can trigger hundreds of runs.
•Essentially: structured, targeted scaling instead of brute-force scaling.
A quiet release but a very loud signal that proof-centric AI is accelerating faster than expected.
AI Post 🪙 | Our X 🥇
👍 30❤ 8🔥 1
00:54
Video unavailableShow in Telegram
🇨🇳 In China, parents are turning AI into a real-time study supervisor.
They open ByteDance’s Doubao, point the camera at their child, and issue a simple command: “Doubao, keep an eye on him. Warn him if he loses focus or slouches.”
From there, the assistant takes over offering a steady stream of corrections:
“Stop fiddling with the pen.”
“Focus on the assignment.”
“Sit up straight.”
“Don’t doze off.”
“Don’t lean on your hand.”
“Quit chewing the pen.”
It’s AI as a digital study hall monitor.
AI Post 🪙 | Our X 🥇
IMG_1364.MP44.53 MB
😁 31😨 18👍 11❤ 3🤪 1
