Now you can choose models on OpenRouter based on cache hit rate! Very helpful when budeting token costs. The Amazon of tokens keeps delivering useful features.
openrouter.ai/anthropic/claude-opus-4.8/pricing
openrouter.ai/anthropic/claude-opus-4.8/pricing
👀1
🇨🇭 Global One SI - Prompt
How much of a big deal is Apple going to make of AI Siri next week? Could we see M5 Mac Studio?
APPLE $AAPL HAD A SECRET CRISIS MEETING IN EARLY 2025 ABOUT AI
The topic: Apple Intelligence was a flop. The new Siri was about to be delayed. And rivals like OpenAI, Google, and Meta were lapping them.
Here's what came out of that meeting:
John Giannandrea, Apple's AI chief, was stripped of his portfolio and pushed out. Mike Rockwell, the engineer behind Vision Pro, was handed Siri. A deal was struck with Google to replace the underlying Siri models with Gemini and Google Cloud.
Tim Cook himself became more involved in AI than any other product initiative in a decade, personally voicing feature preferences and overriding his own lieutenants.
This Monday at WWDC, the result of all of it goes public: a completely rebuilt Siri with a standalone ChatGPT-rival app.
The new Siri is also one of Cook's final acts as CEO before John Ternus takes over
The topic: Apple Intelligence was a flop. The new Siri was about to be delayed. And rivals like OpenAI, Google, and Meta were lapping them.
Here's what came out of that meeting:
John Giannandrea, Apple's AI chief, was stripped of his portfolio and pushed out. Mike Rockwell, the engineer behind Vision Pro, was handed Siri. A deal was struck with Google to replace the underlying Siri models with Gemini and Google Cloud.
Tim Cook himself became more involved in AI than any other product initiative in a decade, personally voicing feature preferences and overriding his own lieutenants.
This Monday at WWDC, the result of all of it goes public: a completely rebuilt Siri with a standalone ChatGPT-rival app.
The new Siri is also one of Cook's final acts as CEO before John Ternus takes over
🇨🇭 Global One SI - Prompt
The rumored ChatGPT superapp could be huge. From what’s being reported, OpenAI may be moving toward one unified app that brings together: • ChatGPT for thinking, writing, planning, research and everyday AI assistance • Codex for coding, debugging, building…
GPT 5.6 looks very close.
Nothing official yet, but the rumors are getting spicy:
• much stronger reasoning
• massive long context, possibly around 1.5M tokens
• better coding and full codebase understanding
• deeper Codex integration
• stronger agentic workflows
• faster tool use across apps
• better multimodal work
• and the big one: computer use pushing above human level on real desktop tasks
If even half of this is true, GPT 5.6 will be another step toward AI that can actually operate software, finish work, and move through the digital world like a real agent.
Nothing official yet, but the rumors are getting spicy:
• much stronger reasoning
• massive long context, possibly around 1.5M tokens
• better coding and full codebase understanding
• deeper Codex integration
• stronger agentic workflows
• faster tool use across apps
• better multimodal work
• and the big one: computer use pushing above human level on real desktop tasks
If even half of this is true, GPT 5.6 will be another step toward AI that can actually operate software, finish work, and move through the digital world like a real agent.
🇨🇭 Global One SI - Prompt
🚀 Windows + NVIDIA/GB300 bringing DGX Station to the masses. A deskside AI supercomputer with: ✦ 252 GB HBM3e VRAM ✦ 748 GB total unified memory 💰 At an estimated $80 - $300K, who is this actually for? Enterprise devs? Researchers? Power users? Big AI teams?…
The new Nvidia DGX Station for Windows
The Ultra Desktop comes with up to 784GB of coherent memory and up to 20 petaflops FP4 compute
It can handle up to 1 trillion parameters locally
Systems based on it are available priced at up to around $85,000
The Ultra Desktop comes with up to 784GB of coherent memory and up to 20 petaflops FP4 compute
It can handle up to 1 trillion parameters locally
Systems based on it are available priced at up to around $85,000
🇨🇭 Global One SI - Prompt
The new Nvidia DGX Station for Windows The Ultra Desktop comes with up to 784GB of coherent memory and up to 20 petaflops FP4 compute It can handle up to 1 trillion parameters locally Systems based on it are available priced at up to around $85,000
This media is not supported in your browser
VIEW IN TELEGRAM
*For #Linux. Nobody wants Windows
Chinese startup Monako unveils smart glasses built to run AI coding agents like Claude Code & Codex.
monako.ai
monako.ai
🇨🇭 Global One SI - Prompt
How much of a big deal is Apple going to make of AI Siri next week? Could we see M5 Mac Studio?
This media is not supported in your browser
VIEW IN TELEGRAM
A developer makes $14k/month from 4 Mac Minis.
Together they run 70B+ models locally more than a single Mac Mini could handle.
Most people rent cloud GPUs every month.
He paid once and built his own AI cluster.
Started with one Mac Mini.
Then another.
Then another.
Then another.
Clients pay for inference.
Data never leaves the room.
Scaling just means stacking one more box on the desk.
Most people rent compute.
He owns it.
github.com/exo-explore/exo
Together they run 70B+ models locally more than a single Mac Mini could handle.
Most people rent cloud GPUs every month.
He paid once and built his own AI cluster.
Started with one Mac Mini.
Then another.
Then another.
Then another.
Clients pay for inference.
Data never leaves the room.
Scaling just means stacking one more box on the desk.
Most people rent compute.
He owns it.
github.com/exo-explore/exo
✍1
🇨🇭 Global One SI - Prompt
Photo
This media is not supported in your browser
VIEW IN TELEGRAM
Introducing Monako Glass 👓
The world's first wearable #Linux computer in glasses form.
Run Claude Code, Codex, and any coding agent — anywhere.
The world's first wearable #Linux computer in glasses form.
Run Claude Code, Codex, and any coding agent — anywhere.
🇨🇭 Global One SI - Prompt
Introducing Monako Glass 👓 The world's first wearable #Linux computer in glasses form. Run Claude Code, Codex, and any coding agent — anywhere.
This media is not supported in your browser
VIEW IN TELEGRAM
Vibe coding will never be the same again!
Now your Claude Code is literally sitting in your chair looking at your screen :)
#Monako #AICoding #AIGlasses #ARGlass #BuildingInPublic #DevTools #VibeCoding #CreativeTools #Productivity
Now your Claude Code is literally sitting in your chair looking at your screen :)
#Monako #AICoding #AIGlasses #ARGlass #BuildingInPublic #DevTools #VibeCoding #CreativeTools #Productivity
✍1
$HYLN - Why invest
1. Massive Market Pivot (AI Data Centers)
Hyliion successfully shifted from a struggling EV-truck maker to a clean-tech energy provider. It now perfectly targets the global power shortage faced by AI data centers, which desperately need independent, on-site electricity.
2. Revolutionary KARNO Technology
The KARNO generator is a massive upgrade over traditional diesel generators:
Fuel-Agnostic: Seamlessly switches between hydrogen, natural gas, biogas, and diesel with zero hardware changes.
Ultra-Low Maintenance: Built via 3D printing with only one moving part per shaft. It requires no oil, no spark plugs, and no regular service.
Efficient Architecture: Delivers direct 800V DC power, eliminating energy losses during conversion for data centers.
3. Top-Tier Validation & Military backing
This is no longer a concept stage. The technology is heavily backed and validated by:
U.S. Navy contracts for stealth, low-signature maritime drones.
ABM Industries for nationwide commercial distribution.
4. Explosive Growth Momentum
Driven by surging demand and a massive earnings beat in early 2026, the stock has shown incredible market momentum, more than tripling in price within just a few months.
1. Massive Market Pivot (AI Data Centers)
Hyliion successfully shifted from a struggling EV-truck maker to a clean-tech energy provider. It now perfectly targets the global power shortage faced by AI data centers, which desperately need independent, on-site electricity.
2. Revolutionary KARNO Technology
The KARNO generator is a massive upgrade over traditional diesel generators:
Fuel-Agnostic: Seamlessly switches between hydrogen, natural gas, biogas, and diesel with zero hardware changes.
Ultra-Low Maintenance: Built via 3D printing with only one moving part per shaft. It requires no oil, no spark plugs, and no regular service.
Efficient Architecture: Delivers direct 800V DC power, eliminating energy losses during conversion for data centers.
3. Top-Tier Validation & Military backing
This is no longer a concept stage. The technology is heavily backed and validated by:
U.S. Navy contracts for stealth, low-signature maritime drones.
ABM Industries for nationwide commercial distribution.
4. Explosive Growth Momentum
Driven by surging demand and a massive earnings beat in early 2026, the stock has shown incredible market momentum, more than tripling in price within just a few months.
🇨🇭 Global One SI - Prompt
$LASE - Did you listen?
$LASE continues to strengthen its position in both the industrial and defense sectors through a combination of high-profile customer relationships and recent government recognition.
The company has highlighted work with globally recognized organizations and industry leaders, including Coca-Cola, General Electric (GE), Harley-Davidson, NASA, Eaton, Boeing, Lockheed Martin, and Honeywell.
These relationships span aerospace, defense, manufacturing, infrastructure, and industrial markets, providing strong validation for Laser Photonics' technology and demonstrating its ability to serve demanding customers across multiple industries.
A major recent development came on June 2, 2026, when Laser Photonics announced that its Laser Shield Anti-Drone (LSAD) system had been selected by the U.S. Department of War under the Mission Engineering and Integration Activity (MEIA) Vulcan Call for Solutions program. The LSAD system was recognized as one of the top submissions in the Counter C5ISR-T category following a rigorous evaluation by government mission engineering and technical teams. As a result, the company was invited to participate in an exclusive technical exchange with government engineers, further validating the effectiveness and potential of its counter-drone technology.
This achievement represents an important milestone for Laser Photonics as it expands beyond its established industrial laser solutions into the rapidly growing defense and national security market. The recognition from a U.S. government program highlights the increasing relevance of directed-energy and counter-drone technologies, areas that are expected to see significant investment in the coming years. While this selection is not yet a major procurement contract, it can serve as a gateway to future testing programs, pilot deployments, and potential government purchasing opportunities.
For investors, the combination of blue-chip customers, expanding defense capabilities, and growing government interest in the company's Laser Shield Anti-Drone platform positions Laser Photonics as an emerging player in both advanced manufacturing and next-generation defense technologies. The company's ability to secure recognition from government agencies while maintaining relationships with some of the world's most respected industrial and aerospace organizations could significantly enhance its visibility and long-term growth prospects.
The company has highlighted work with globally recognized organizations and industry leaders, including Coca-Cola, General Electric (GE), Harley-Davidson, NASA, Eaton, Boeing, Lockheed Martin, and Honeywell.
These relationships span aerospace, defense, manufacturing, infrastructure, and industrial markets, providing strong validation for Laser Photonics' technology and demonstrating its ability to serve demanding customers across multiple industries.
A major recent development came on June 2, 2026, when Laser Photonics announced that its Laser Shield Anti-Drone (LSAD) system had been selected by the U.S. Department of War under the Mission Engineering and Integration Activity (MEIA) Vulcan Call for Solutions program. The LSAD system was recognized as one of the top submissions in the Counter C5ISR-T category following a rigorous evaluation by government mission engineering and technical teams. As a result, the company was invited to participate in an exclusive technical exchange with government engineers, further validating the effectiveness and potential of its counter-drone technology.
This achievement represents an important milestone for Laser Photonics as it expands beyond its established industrial laser solutions into the rapidly growing defense and national security market. The recognition from a U.S. government program highlights the increasing relevance of directed-energy and counter-drone technologies, areas that are expected to see significant investment in the coming years. While this selection is not yet a major procurement contract, it can serve as a gateway to future testing programs, pilot deployments, and potential government purchasing opportunities.
For investors, the combination of blue-chip customers, expanding defense capabilities, and growing government interest in the company's Laser Shield Anti-Drone platform positions Laser Photonics as an emerging player in both advanced manufacturing and next-generation defense technologies. The company's ability to secure recognition from government agencies while maintaining relationships with some of the world's most respected industrial and aerospace organizations could significantly enhance its visibility and long-term growth prospects.
👀1
🇨🇭 Global One SI - Prompt
$BB Everyone loved BlackBerry at $10. Suddenly they’re worried at $9. Nothing has changed in my thesis. QNX is quietly becoming part of the infrastructure layer powering AI in the physical world. Cars. Robotics. Industrial automation. Healthcare. Aerospace.…
The Quiet Expansion of QNX: How BlackBerry Is Building the Operating System for the Physical AI Economy
Most investors still think of QNX ( $BB ) as an automotive story.
That perception is understandable. For years, QNX's real-time operating system (RTOS) has been one of the hidden foundations of modern vehicles, powering infotainment systems, digital cockpits, advanced driver assistance systems (ADAS), and other safety-critical functions. Today, QNX software is deployed in hundreds of millions of vehicles worldwide and is used by many of the world's largest automakers.
Yet a closer examination of QNX's recent job postings reveals something much bigger taking shape.
Taken together, the hiring activity suggests that QNX is quietly repositioning itself from an automotive software company into a foundational platform for what it increasingly describes as "physical AI"—the emerging world of intelligent machines operating in the real world.
This is not a flashy consumer-facing strategy. There are no viral chatbots or consumer gadgets involved. Instead, QNX appears to be methodically expanding into industries where software failures can have real-world consequences: robotics, industrial automation, medical devices, aerospace, rail, commercial transportation, and defense.
The result is a picture of a company attempting to build a decades-long moat around some of the most demanding computing environments on Earth.
Beyond Automotive: The Great Diversification
One of the most striking themes in the job postings is how often non-automotive markets appear.
Recent openings include:
- Senior Technical Product Manager (Industrial Automation & Robotics)
- Field Application Engineers supporting industrial automation, robotics, and medical devices
- Senior Product & Industry Marketing Manager (Robotics & Embedded Systems)
- Cybersecurity leadership roles spanning multiple regulated industries
These positions indicate that QNX is actively pursuing markets beyond its traditional automotive stronghold.
This is significant because many of these industries face a similar challenge: integrating increasingly sophisticated AI and automation systems into environments where reliability and safety are paramount.
Factories deploying autonomous robots.
Hospitals using intelligent medical equipment.
Rail networks relying on real-time signaling systems.
Defense and aerospace platforms requiring deterministic performance.
These are not markets where software crashes can simply be tolerated.
They are precisely the types of environments where QNX's deterministic microkernel architecture and safety certifications provide a meaningful advantage.
The Rise of Physical AI
Perhaps the most revealing language appears in QNX's own descriptions of its mission.
Rather than describing itself solely as an automotive software company, QNX increasingly refers to providing the trusted foundation for "software-defined systems" and "physical AI systems."
That phrase deserves attention.
The AI industry has spent the last several years focused primarily on digital intelligence: chatbots, search engines, content generation, and virtual assistants.
The next phase may involve AI moving into the physical world.
Robots.
Industrial machines.
Autonomous systems.
Medical devices.
Transportation infrastructure.
In these environments, intelligence alone is insufficient.
A robot must move safely.
A train must stop predictably.
A medical device must operate reliably.
An autonomous machine must respond within precise timing constraints.
The challenge shifts from generating text to controlling matter.
That transition greatly increases the importance of deterministic operating systems.
QNX appears to be positioning itself as one of the software foundations upon which physical AI systems will be built.
Building the Layer Beneath NVIDIA
One of the most overlooked clues appears in a Senior Technical Product Manager posting.
Most investors still think of QNX ( $BB ) as an automotive story.
That perception is understandable. For years, QNX's real-time operating system (RTOS) has been one of the hidden foundations of modern vehicles, powering infotainment systems, digital cockpits, advanced driver assistance systems (ADAS), and other safety-critical functions. Today, QNX software is deployed in hundreds of millions of vehicles worldwide and is used by many of the world's largest automakers.
Yet a closer examination of QNX's recent job postings reveals something much bigger taking shape.
Taken together, the hiring activity suggests that QNX is quietly repositioning itself from an automotive software company into a foundational platform for what it increasingly describes as "physical AI"—the emerging world of intelligent machines operating in the real world.
This is not a flashy consumer-facing strategy. There are no viral chatbots or consumer gadgets involved. Instead, QNX appears to be methodically expanding into industries where software failures can have real-world consequences: robotics, industrial automation, medical devices, aerospace, rail, commercial transportation, and defense.
The result is a picture of a company attempting to build a decades-long moat around some of the most demanding computing environments on Earth.
Beyond Automotive: The Great Diversification
One of the most striking themes in the job postings is how often non-automotive markets appear.
Recent openings include:
- Senior Technical Product Manager (Industrial Automation & Robotics)
- Field Application Engineers supporting industrial automation, robotics, and medical devices
- Senior Product & Industry Marketing Manager (Robotics & Embedded Systems)
- Cybersecurity leadership roles spanning multiple regulated industries
These positions indicate that QNX is actively pursuing markets beyond its traditional automotive stronghold.
This is significant because many of these industries face a similar challenge: integrating increasingly sophisticated AI and automation systems into environments where reliability and safety are paramount.
Factories deploying autonomous robots.
Hospitals using intelligent medical equipment.
Rail networks relying on real-time signaling systems.
Defense and aerospace platforms requiring deterministic performance.
These are not markets where software crashes can simply be tolerated.
They are precisely the types of environments where QNX's deterministic microkernel architecture and safety certifications provide a meaningful advantage.
The Rise of Physical AI
Perhaps the most revealing language appears in QNX's own descriptions of its mission.
Rather than describing itself solely as an automotive software company, QNX increasingly refers to providing the trusted foundation for "software-defined systems" and "physical AI systems."
That phrase deserves attention.
The AI industry has spent the last several years focused primarily on digital intelligence: chatbots, search engines, content generation, and virtual assistants.
The next phase may involve AI moving into the physical world.
Robots.
Industrial machines.
Autonomous systems.
Medical devices.
Transportation infrastructure.
In these environments, intelligence alone is insufficient.
A robot must move safely.
A train must stop predictably.
A medical device must operate reliably.
An autonomous machine must respond within precise timing constraints.
The challenge shifts from generating text to controlling matter.
That transition greatly increases the importance of deterministic operating systems.
QNX appears to be positioning itself as one of the software foundations upon which physical AI systems will be built.
Building the Layer Beneath NVIDIA
One of the most overlooked clues appears in a Senior Technical Product Manager posting.
👀1
‼️The AI data center buildout is generating more headlines than capacity:
More than 60% of US data center capacity planned for completion in 2027 IS NOT YET under construction, and another 7% is already delayed, according to a JPMorgan analysis.
Of the ~35 gigawatts of data center capacity with unknown completion dates, only a fraction is actually under construction, with the vast majority remaining at the announcement stage.
This comes as supply chain backlogs, permitting fights, power grid connection delays, and shortages of gas turbines and electrical transformers are all contributing to a widening gap between announced ambitions and actual progress.
Microsoft, Alphabet, Meta, and Amazon collectively spent $410 billion on capital expenditure last year and are expected to spend more than $670 billion in 2026, yet the physical infrastructure to support this spending is falling significantly behind schedule.
Alphabet highlighted the scale of this challenge by raising $80 billion in new equity this week, its largest ever capital raise, prompting its stock to fall as investors questioned the true extent of AI capital spending needs.
Investors are beginning to question whether the AI capital spending cycle can deliver returns fast enough to justify the cost.
More than 60% of US data center capacity planned for completion in 2027 IS NOT YET under construction, and another 7% is already delayed, according to a JPMorgan analysis.
Of the ~35 gigawatts of data center capacity with unknown completion dates, only a fraction is actually under construction, with the vast majority remaining at the announcement stage.
This comes as supply chain backlogs, permitting fights, power grid connection delays, and shortages of gas turbines and electrical transformers are all contributing to a widening gap between announced ambitions and actual progress.
Microsoft, Alphabet, Meta, and Amazon collectively spent $410 billion on capital expenditure last year and are expected to spend more than $670 billion in 2026, yet the physical infrastructure to support this spending is falling significantly behind schedule.
Alphabet highlighted the scale of this challenge by raising $80 billion in new equity this week, its largest ever capital raise, prompting its stock to fall as investors questioned the true extent of AI capital spending needs.
Investors are beginning to question whether the AI capital spending cycle can deliver returns fast enough to justify the cost.