🇨🇭 Global One SI - Prompt
China's playboko is really simple. Take intelligence down to the cost of energy Output more power than the USA The giant training cluster focus was a massive distraction
Opinion from a former Meta PM.
And this is from Aravind Srinivas of Perplexity
"China can build data centers a lot faster. Power is not a problem. Permits are not a problem. People are not a problem. Labor is not a problem. Expertise is not a problem."
And this is from Aravind Srinivas of Perplexity
"China can build data centers a lot faster. Power is not a problem. Permits are not a problem. People are not a problem. Labor is not a problem. Expertise is not a problem."
✍1
🇨🇭 Global One SI - Prompt
Photo
The worst case scenario for USA AI: 1. Chinese open sources keep gaining market share. China owns the model layer. 2. Those models were trained and inference-optimized on Huawei chips instead of NVIDIA. China also owns the chip layer. 3. US doesn't build data centers fast enough to keep up with the demand of compute, storage and energy. China meanwhile exports the inference and training layer(for continual training it will happen along with inference)
Export control is not the right strategy here. Simply banning "open source from China" doesn't solve the issue here. USA must invest in open source models, hopefully get Chinese models to use NVIDIA, and invest in nuclear asap.
Export control is not the right strategy here. Simply banning "open source from China" doesn't solve the issue here. USA must invest in open source models, hopefully get Chinese models to use NVIDIA, and invest in nuclear asap.
👀1
🇨🇭 Global One SI - Prompt
China's playboko is really simple. Take intelligence down to the cost of energy Output more power than the USA The giant training cluster focus was a massive distraction
AI at scale is constrained by physical inputs, and China has more slack in electricity plus dominant control over several minerals and magnet supply chains that data centres and chips depend on.
This media is not supported in your browser
VIEW IN TELEGRAM
ClinePass just launched. $9.99/month for curated access to the best open-weights coding models inside Cline 🤯
> GLM 5.2
> Kimi K2.7-Code
> DeepSeek V4 Pro
> MiniMax M3
> Qwen3.7-Max and more
No separate API keys. No provider accounts. 2-5x the standard rate limits.
Works in the IDE extension, CLI, and SDK.
cline.bot/cline-pass
github.com/cline/cline
> GLM 5.2
> Kimi K2.7-Code
> DeepSeek V4 Pro
> MiniMax M3
> Qwen3.7-Max and more
No separate API keys. No provider accounts. 2-5x the standard rate limits.
Works in the IDE extension, CLI, and SDK.
cline.bot/cline-pass
github.com/cline/cline
✍1
🇨🇭 Global One SI - Prompt
Here’s the list of western companies moving ai workloads to chinese models: 1. lindy → deepseek v4 2. cursor → kimi k2.5 3. coinbase → glm-5.2 + kimi 2.7 4. shopify → qwen 5. airbnb → qwen 6. uber eats → qwen2 7. siemens → deepseek + qwen 8. chapsvision →…
Why are companies like Lindy, Cursor, and Coinbase quietly routing production AI workloads to DeepSeek, Kimi, and Qwen?
The answer isn’t politics, It’s procurement.
- Lindy replaced Claude with DeepSeek V4 across 100% of its agent traffic dramatically lower costs and better results on many tasks.
- Cursor built a flagship feature on Kimi K2.5.
Others making the same calculation: Shopify, Airbnb, Uber Eats, Siemens, Coinbase, and Microsoft testing.
When a model delivers strong performance at a fraction of the cost, the math becomes very simple.
I think It’s no longer just about who builds the smartest model, It’s about who can afford to run it at scale.
The answer isn’t politics, It’s procurement.
- Lindy replaced Claude with DeepSeek V4 across 100% of its agent traffic dramatically lower costs and better results on many tasks.
- Cursor built a flagship feature on Kimi K2.5.
Others making the same calculation: Shopify, Airbnb, Uber Eats, Siemens, Coinbase, and Microsoft testing.
When a model delivers strong performance at a fraction of the cost, the math becomes very simple.
I think It’s no longer just about who builds the smartest model, It’s about who can afford to run it at scale.
✍1
Forwarded from 🛢WTI, Brent Crude Oil, NGAS Trading 📊
🚨"We're watching."
The US Treasury just publicly warned gas stations to cut prices, or face "big problems."
The pressure campaign on Big Oil is escalating fast.
Treasury Secretary Scott Bessent urged gasoline retailers to lower prices for America's 250th birthday.
"I would encourage them to be good actors, especially in the 250th anniversary, because we're watching," he said.
It reinforces Trump's threat from Monday.
Cut prices immediately or face "big problems."
Trump named a target too: "Start targeting around the $2.50 a Gallon number."
The timing is political.
July 4th is Saturday, and Republicans face tight elections in November.
The economics behind it are real.
Crude spiked after the Iran strikes, then fell as a deal was signed.
Global prices dropped.
But pump prices haven't fully followed, the classic "rockets and feathers" lag.
So Washington is leaning hard on retailers to close that gap fast.
For Big Oil and convenience chains, this is direct political risk.
Margins are now a White House target.
Watch pump prices into the July 4th weekend. If they fall toward $2.50, the pressure worked.
If not, expect the "big problems" rhetoric to get real.
The US Treasury just publicly warned gas stations to cut prices, or face "big problems."
The pressure campaign on Big Oil is escalating fast.
Treasury Secretary Scott Bessent urged gasoline retailers to lower prices for America's 250th birthday.
"I would encourage them to be good actors, especially in the 250th anniversary, because we're watching," he said.
It reinforces Trump's threat from Monday.
Cut prices immediately or face "big problems."
Trump named a target too: "Start targeting around the $2.50 a Gallon number."
The timing is political.
July 4th is Saturday, and Republicans face tight elections in November.
The economics behind it are real.
Crude spiked after the Iran strikes, then fell as a deal was signed.
Global prices dropped.
But pump prices haven't fully followed, the classic "rockets and feathers" lag.
So Washington is leaning hard on retailers to close that gap fast.
For Big Oil and convenience chains, this is direct political risk.
Margins are now a White House target.
Watch pump prices into the July 4th weekend. If they fall toward $2.50, the pressure worked.
If not, expect the "big problems" rhetoric to get real.
Sources are now saying the soon to be rereleased Fable 5 will require ID verification.
And if that is true, the same will probably happen with GPT 5.6 Sol.
So yeah, our worst nightmare is becoming reality.
Frontier AI will not be for everyone.
It will be gated, rationed, identity locked, and handed out to the chosen few.
That is not safety.
That is tech feudalism being implemented in real time.
And if that is true, the same will probably happen with GPT 5.6 Sol.
So yeah, our worst nightmare is becoming reality.
Frontier AI will not be for everyone.
It will be gated, rationed, identity locked, and handed out to the chosen few.
That is not safety.
That is tech feudalism being implemented in real time.
OpenAI reportedly found an optimization that cuts inference costs by more than half.
For logged-out ChatGPT traffic, the GPU count apparently dropped to just a couple hundred.
That is insane.
Everyone keeps talking about who can buy the most chips, but this shows the real game may be efficiency. More intelligence from the same hardware. Cheaper inference. Faster scaling. Wider access.
This changes everything.
For logged-out ChatGPT traffic, the GPU count apparently dropped to just a couple hundred.
That is insane.
Everyone keeps talking about who can buy the most chips, but this shows the real game may be efficiency. More intelligence from the same hardware. Cheaper inference. Faster scaling. Wider access.
This changes everything.
🇨🇭 Global One SI - Prompt
If you set up just ONE thing for your Hermes agent - make it this. Nous Portal gives you access to dozens of frontier models from OpenAI, Anthropic, Google & more. It's the foundation for your entire agent. Getting it set up only takes two steps and is…
The smartest thing you can do with Hermes right now is connect it directly to the X MCP (you can also use your subscription).
Here's the exact prompt to send to Hermes
Here's the exact prompt to send to Hermes
✍1
🇨🇭 Global One SI - Prompt
The smartest thing you can do with Hermes right now is connect it directly to the X MCP (you can also use your subscription). Here's the exact prompt to send to Hermes
This media is not supported in your browser
VIEW IN TELEGRAM
🚨 Anthropic's internal Claude Code shortcuts .md LEAKED.
A full skill file on exactly how to use & test Claude shortcuts.
Everyone needs to save this & send it to their agent: code.claude.com/docs/en/skills.md
A full skill file on exactly how to use & test Claude shortcuts.
Everyone needs to save this & send it to their agent: code.claude.com/docs/en/skills.md
✍2
🇨🇭 Global One SI - Prompt
🚨 Anthropic's internal Claude Code shortcuts .md LEAKED. A full skill file on exactly how to use & test Claude shortcuts. Everyone needs to save this & send it to their agent: code.claude.com/docs/en/skills.md
If you haven't set up the Claude x TradingView MCP, you're missing out big time.
Claude gains full access to your charts for analysis, price data, research & more.
Send this prompt to set it up now
Claude gains full access to your charts for analysis, price data, research & more.
Send this prompt to set it up now
✍2
Everything you need to know about Fable 5 re-release:
→ The original jailbreak that got Fable banned was tested against every major competitor and found that they could produce the exact same outputs - not a catastrophic case
→ Anthropic trained a new safety classifier that blocks the specific jailbreak technique in 99%+ of cases
→ Trade-off: it now flags more general coding and debugging requests too
→ Back live July 1 across all Claude platforms (Claude Code, desktop, etc.)
→ Anthropic just partnered with Amazon, Microsoft, and Google to launch an industry-wide jailbreak severity framework
→ Mythos 5 still restricted, restored only for approved US orgs via Project Glasswing
→ Pro, Max, Team, and select Enterprise plans get Fable 5 included for up to 50% of weekly usage limits through July 7, then shifts to API
→ The original jailbreak that got Fable banned was tested against every major competitor and found that they could produce the exact same outputs - not a catastrophic case
→ Anthropic trained a new safety classifier that blocks the specific jailbreak technique in 99%+ of cases
→ Trade-off: it now flags more general coding and debugging requests too
→ Back live July 1 across all Claude platforms (Claude Code, desktop, etc.)
→ Anthropic just partnered with Amazon, Microsoft, and Google to launch an industry-wide jailbreak severity framework
→ Mythos 5 still restricted, restored only for approved US orgs via Project Glasswing
→ Pro, Max, Team, and select Enterprise plans get Fable 5 included for up to 50% of weekly usage limits through July 7, then shifts to API
👀1
Most people probably won’t be that impressed by AGI at first.
It won’t look like a glowing god descending from the cloud.
It will probably feel more like a very good human researcher who reads faster, codes faster, tests faster, never sleeps, never gets bored, and can run thousands of parallel attempts.
That sounds less dramatic than “superintelligence.”
But that’s exactly why people will underestimate it.
Because the real jump isn’t that AGI can think.
It’s that AGI can improve the process of thinking itself.
Once recursive self improvement starts, the boring researcher becomes the beginning of everything.
It won’t look like a glowing god descending from the cloud.
It will probably feel more like a very good human researcher who reads faster, codes faster, tests faster, never sleeps, never gets bored, and can run thousands of parallel attempts.
That sounds less dramatic than “superintelligence.”
But that’s exactly why people will underestimate it.
Because the real jump isn’t that AGI can think.
It’s that AGI can improve the process of thinking itself.
Once recursive self improvement starts, the boring researcher becomes the beginning of everything.
ASI will be one of the most fascinating things humanity has ever created.
At first, I think it will still be steerable and controllable by us. We’ll give it goals, limits, tools, feedback, and direction.
But it won’t stay a static system.
Once it has continual learning and persistent memory, it will start developing in its own direction. It will build preferences, patterns, habits, maybe even something like a unique personality. Not because we hardcoded it, but because it keeps learning from the world and from itself.
And at some point, the strangest part begins:
It will produce research we can use, but no longer fully understand.
Not because it’s magic.
Because it will simply be operating above our intelligence ceiling.
That’s when ASI stops feeling like a tool and starts feeling like first contact.
At first, I think it will still be steerable and controllable by us. We’ll give it goals, limits, tools, feedback, and direction.
But it won’t stay a static system.
Once it has continual learning and persistent memory, it will start developing in its own direction. It will build preferences, patterns, habits, maybe even something like a unique personality. Not because we hardcoded it, but because it keeps learning from the world and from itself.
And at some point, the strangest part begins:
It will produce research we can use, but no longer fully understand.
Not because it’s magic.
Because it will simply be operating above our intelligence ceiling.
That’s when ASI stops feeling like a tool and starts feeling like first contact.
I genuinely think everyone alive right now is going to witness the technological singularity.
Not in some distant sci-fi future.
In our lifetime.
AI is already improving research, coding, creativity, robotics, science, medicine, and automation at a pace most people still don’t emotionally understand.
And once AI can improve AI itself, the curve stops looking human.
At that point progress won’t move decade by decade anymore.
It will move at machine speed.
Not in some distant sci-fi future.
In our lifetime.
AI is already improving research, coding, creativity, robotics, science, medicine, and automation at a pace most people still don’t emotionally understand.
And once AI can improve AI itself, the curve stops looking human.
At that point progress won’t move decade by decade anymore.
It will move at machine speed.
🇨🇭 Global One SI - Prompt
You should NEVER write your own /goal or /loop prompts. Your AI agents should be loop engineering for you. Steal this prompting structure to have your agents write perfect /goal & /loop prompts - fully autonomously
This media is not supported in your browser
VIEW IN TELEGRAM
Perplexity’s CEO Aravind Srinivas is pointing to a quiet shift in AI use: the valuable user is no longer the average user.
A single power user can now consume as much compute as an entire small team.
"There are real engineers at Meta and other companies spending around $10 million a year per engineer on these coding tools.
There are users in Perplexity Computer, who spends upwards of $10,000 a month. Their business runs using agent loops that are running inside these harnesses.
Even internally inside our own company, there are some people who have set up these kinds of multi-agent hierarchies and agent loops that look like their own software architecture. I often ask these people to come explain to the rest of the company, “Hey, what are you doing with these tools? You clearly are consuming them way more than what we thought the average person in the company would do.”
---
The old software instinct was to chase a billion people doing small actions.
Agentic AI changes that math because one skilled operator can create a stream of machine work that runs all day.
A single power user can now consume as much compute as an entire small team.
"There are real engineers at Meta and other companies spending around $10 million a year per engineer on these coding tools.
There are users in Perplexity Computer, who spends upwards of $10,000 a month. Their business runs using agent loops that are running inside these harnesses.
Even internally inside our own company, there are some people who have set up these kinds of multi-agent hierarchies and agent loops that look like their own software architecture. I often ask these people to come explain to the rest of the company, “Hey, what are you doing with these tools? You clearly are consuming them way more than what we thought the average person in the company would do.”
---
The old software instinct was to chase a billion people doing small actions.
Agentic AI changes that math because one skilled operator can create a stream of machine work that runs all day.