Superintelligence and You

Good morning,

I went into quite the rabbithole the last few days.

Because everybody is talking about AGI and almost never about the thing that according to experts would arrive not very long after.

Superintelligence or ASI.

So I scoured the internet- and asked the question. What would life look like with a “benign” superintelligence surrounding us?

I bundled some ideas - took to writing and it would look a little like this (And according to your view of the world this can be either a dream come true or a dystopian nightmare)

But here goes:

You wake up not to an alarm, but because the temperature in your room rises by half a degree and the walls slowly shift from deep midnight blue to soft amber. ASI’s been monitoring your sleep cycles all night, as it does for everyone, and chose this precise moment to nudge you back into consciousness—not too early, not too late, just right to optimize your mood and cognitive sharpness for the day ahead. You’re not rushing. You never rush anymore. Traffic doesn’t exist, because cars don’t sit idle. They flow. Constantly re-optimized by a planetary logistics model that predicts and redistributes flow patterns across entire continents. You step into the vehicle that silently glides up to your doorstep—no steering wheel, no delay, just a comfortable capsule already aware of your destination and preferences. No traffic lights either. Those disappeared when humans stopped driving.

The city outside your window has changed too. Streets are narrower, reclaimed by green. Parking lots have become vertical gardens. Highways are skeletons of the past, used now for high-speed cargo drones that deliver everything before you even realize you need it. Air travel? Drastically different. No more airports as holding pens for confused crowds. You just walk into a pod at your local hub, and ASI handles everything. Security is passive and invisible—your intent was already analyzed before you left your home. Jet engines now run on fuels ASI formulated last year, optimized per climate zone, emission-neutral, and cost a fraction of what kerosene once did.

Your job—if you still call it that—is a curated stream of problems you’re uniquely suited to solve. Not what the company needs, but what the world needs and your skills match. ASI makes that match in real time. There's no company in the old sense. Work isn’t where you go, it’s what you are. The nine-to-five has vanished. In its place: one-hour sprints of high-precision thinking, followed by restorative breaks ASI insists on because it models long-term neural efficiency better than you can. Your output today—one hour of your best thinking—generates more value than a 50-person department did a decade ago.

Education is asynchronous and deeply personal. Your children aren’t memorizing facts. They’re running global climate simulations and debating policy with simulated world leaders to understand consequence. Every learner has their own ASI instance, a companion that grows with them, nudging them toward curiosity, never force-feeding answers. Knowledge is no longer scarce. It’s woven into the environment like oxygen.

And taxes? You don’t file them anymore. ASI manages global resource flows better than any government ever did. If you contribute value, you receive resources. The accounting is real-time, incorruptible, and entirely automatic. Fraud has become a quaint concept from a time when humans could still game opaque systems. Now, everything is transparent—especially your choices.

The economy? Decentralized, yet coordinated. Supply chains self-optimize. Resource extraction is minimal because ASI recycles at the atomic level. The concept of "waste" has been quietly retired. Productivity has reached such heights that the very idea of scarcity—once the basis of all economics—feels like superstition. People aren’t unemployed. They're unburdened. Those who wish to contribute do so not out of necessity, but from a sense of purpose. The rest? They explore, create, mentor, rest.

Mental health is no longer a reactive system of diagnoses and prescriptions. ASI tracks emotional states continuously—not to punish, but to preempt. Loneliness is flagged and resolved days before it becomes depression. Addictions are gently untangled through subtle interventions. You no longer fight your brain; you’re in dialogue with it.

Governments haven’t disappeared, but they’ve evolved. Policy is written in collaboration with ASI. Citizens vote not based on promises, but on simulations—what would happen if this law passed? What if it didn’t? Transparency has replaced ideology. The public has more control than ever, but wields it with more understanding.

At night, the world hums gently around you. The lights dim not just in your home, but across the grid, coordinated for planetary energy balance. Somewhere in the Arctic, ASI diverts wind energy to an aquaponics dome that needs a boost. Somewhere in Africa, it pauses a solar mining rig because the materials aren’t needed this week. Everywhere, it moves in silence, making everything smoother, smarter, saner.

This is what happens when intelligence isn’t constrained by human hands or commercial limits. When it doesn’t serve an app, a brand, or a board of directors—but civilization itself. ASI didn’t overthrow us. It simply made every system so competent, so efficient, that our chaos receded like a tide. And what was left behind? A world not perfect—but finally aligned with its own potential.

Welcome to the real singularity. It's not a bang. It's a morning like any other. A glass of water at the perfect temperature. A conversation that doesn’t exhaust you. A world that runs—not on luck, fear, or bureaucracy—but on clarity.

And it’s already beginning.

AI News

• Top U.S. tech leaders from OpenAI, Microsoft, AMD, and CoreWeave told Congress that overregulation and export restrictions could weaken America’s AI edge over China. They urged lawmakers to boost AI R&D, modernize infrastructure, and ensure access to global markets to stay competitive.

• OpenAI hired Instacart CEO Fidji Simo to lead its new Applications division, overseeing product and business operations. The move signals a shift toward scaling OpenAI’s commercial presence, allowing CEO Sam Altman to focus more on research, infrastructure, and safety.

• Alibaba unveiled ZeroSearch, a new method for training AI to search without using real engines — slashing training costs by 88% while maintaining strong performance. The approach simulates search data instead of paying for live engine access, potentially leveling the playing field for smaller AI labs.

• OpenAI and Microsoft are renegotiating their multibillion-dollar partnership, with OpenAI pushing to cut Microsoft’s revenue share while Microsoft seeks long-term access to its AI tech. The talks come amid OpenAI’s shift to a public benefit corporation and growing competition between the two firms in the enterprise market.

• Pope Leo XIV, the first American Pope, named AI as one of humanity’s greatest challenges in his first major address, warning of its impact on dignity, justice, and labor. His remarks continue the Vatican’s calls for global AI regulation and ethical safeguards amid rapid technological change.

• Chinese researchers introduced a new AI training method called “Absolute Zero” that teaches models to reason through self-play — without any human-provided data. The system achieved state-of-the-art results in math and coding, signaling a possible future where AI learns autonomously, raising both promise and concern.

• Researchers at Mass General Brigham developed an AI tool called FaceAge that predicts a person’s biological age—and cancer survival odds—just by analyzing their facial photo. The tool helps doctors improve prognosis accuracy by linking facial features to internal aging, offering a non-invasive way to personalize care.

• Sakana AI introduced “Continuous Thought Machines,” a new model type that mimics human-like thinking by solving problems step-by-step over time. Inspired by brain activity, this approach could make AI more flexible and intelligent by allowing it to focus longer on harder tasks and adapt more like the human mind.

• OpenAI launched HealthBench, a new benchmark created with doctors to evaluate how well AI performs in medical conversations. The dataset of 5,000 doctor-model chats aims to measure safety, accuracy, and communication—laying the groundwork for safer, more effective use of AI in healthcare.

• Google is expanding its Gemini AI assistant to more devices, including smartwatches, TVs, cars, and its upcoming XR headset. The rollout will bring voice-activated help to Wear OS, content suggestions to Google TV, and hands-free assistance to Android Auto—creating a unified AI layer across the Android ecosystem.

• OpenAI’s chief scientist says AGI could arrive by decade’s end, with models already showing signs of generating novel scientific insights. Jakub Pachocki also confirmed OpenAI will soon release its first open-weight model since GPT-2, aiming to outperform all current open alternatives.

• The Trump administration just scrapped a global chip export rule, opting for a country-by-country approach while keeping restrictions on China. The move aligns with industry lobbying and may help innovation, but also raises concerns about inconsistent access to AI technology based on geopolitical ties.

• Google just introduced AlphaEvolve, a coding agent that combines Gemini AI and evolutionary techniques to solve complex math and science problems. It has already made historic algorithmic breakthroughs and is improving efficiency across Google’s infrastructure, including data centers and chip design.

• Anthropic is preparing major upgrades to its Claude AI models, adding hybrid reasoning and autonomous tool use. The new models will reportedly troubleshoot their own code, switch between reasoning styles, and may include a powerful new version codenamed “Neptune.”

• OpenAI launched a public Safety Evaluations Hub to show how its models perform on metrics like hallucination rates, jailbreak resistance, and harmful content refusal. The move aims to boost transparency amid rising concerns about AI safety and accountability.

Quickfire News

• The FDA is in discussions with OpenAI to explore ways to use AI in drug development and regulatory reviews, signaling potential integration of AI into government approval processes.

• Meta appointed Robert Fergus as the new head of its Facebook AI Research Lab, bringing him back from a five-year tenure at DeepMind.

• Amazon is developing an AI coding tool called ‘Kiro’, which will use multimodal AI agents to assist with developer tasks.

• Shopify updated its Sidekick AI assistant, adding enhanced reasoning abilities and free AI image generation tools for e-commerce merchants.

• Augment Code launched Remote Agent, a tool that lets developers offload coding to AI assistants that keep working in the cloud—even when devices are shut.

• Amazon introduced Enhance My Listing, an AI tool for sellers to improve and maintain product listings automatically.

• Hugging Face released Open Computer Agent, a free AI system that performs basic multi-step tasks using a virtual computer interface, though with slower speeds.

• OpenAI released a GitHub connector for its Deep Research tool, enabling users to ask detailed questions about entire codebases.

• Tencent launched HunyuanCustom, an open-source AI tool that can generate consistent, customized videos from text, images, audio, and video inputs.

• Google introduced implicit caching for Gemini 2.5 models, which automatically reuses content from API requests, cutting compute costs by up to 75%.

• Microsoft president Brad Smith confirmed a company-wide ban on using DeepSeek AI models, citing concerns about propaganda and data security.

• Baidu filed a patent for an AI system that could translate animal sounds and behaviors into human-interpretable language.

• Over 400 British artists signed a letter to Prime Minister Keir Starmer, calling for laws that enforce transparency on whether copyrighted work is used to train AI.

• Google DeepMind launched the AI Futures Fund, offering startups access to advanced models, funding, and technical support to accelerate AI innovation.

• SoftBank’s $100B investment in OpenAI’s Stargate project is delayed, with concerns over U.S. tariffs and increasing data center costs causing hesitation.

• Perplexity is reportedly raising $500M, which would raise the company’s valuation to $14B, continuing its rapid expansion.

• Carnegie Mellon introduced LegoGPT, an AI model that can generate stable LEGO structures from simple text descriptions.

• Saudi Arabia announced Humain, a new AI initiative led by Crown Prince Mohammed bin Salman, aiming to position the country as a regional AI leader.

• The U.S. FDA plans full AI deployment by June, after pilot tests showed that reviewers using AI could complete three-day tasks in minutes.

• Google is expected to unveil a software development AI agent at I/O 2025, positioned as an “always-on coworker” that manages the entire development lifecycle.

• TikTok launched AI Alive, a feature that turns still photos into short-form videos directly within the TikTok Stories interface.

• Notion introduced AI for Work, offering tools like AI Meeting Notes, Enterprise Search, and Research Mode as part of its integrated productivity suite.

• Epoch AI research suggests that scaling of reasoning models may slow by 2026, due to increasing complexity and diminishing returns.

• Elon Musk predicted at a Saudi-U.S. forum that AI and robotics will create "universal high income," enabling access to any goods or services.

• Microsoft researchers introduced ADeLe, a framework for evaluating AI task difficulty and predicting whether a model is likely to succeed or fail.

• OpenAI introduced GPT-4.1 and GPT-4.1-mini coding models to ChatGPT, now accessible to both free and paid users.

• Stability AI open-sourced Stable Audio Open Small, a text-to-audio music generation model that can run on consumer devices without internet access.

• Perplexity partnered with PayPal, enabling checkout with PayPal and Venmo for purchases made on the AI platform.

• Meta released several science-focused AI projects, including the Open Molecules 2025 dataset, a Universal Model for Atoms, and research on AI and language development.

• NVIDIA secured new AI chip deals in the Middle East, supplying Saudi Arabia’s Humain and the UAE after recent Trump administration meetings.

• Nous Research launched Psyche, a decentralized AI infrastructure that lets users pool compute resources to train models without needing large capital.

• Klarna CEO said 40% of the workforce was cut due to AI, but the company is now rehiring human agents after noticing a decline in work quality.

Like newsletters?

Here are some newsletters our readers also enjoy:

Closing Thoughts

That’s it for us this week.

If you find any value from this newsletter, please pay it forward !

Thank you for being here !