Usamah Sedqi Al-Baik أسامة صدقي البيك

Artificial intelligence is everywhere—on your phone, in your car, and increasingly inside the tools you use at work. As its presence grows, a familiar question returns: Are there limits to what AI can do? For decades, many confidently predicted hard boundaries. Most of those predictions didn’t age well. Yet, it’s equally true that meaningful limits remain—some technical, some philosophical, and some social. This article organizes the discussion in three parts. First, we clarify how AI fits into the classic data–information–knowledge–wisdom (DIKW) model. Second, we revisit “old limits” that fell and the problems that are only partially solved. Third, we examine the frontiers that still constrain AI and outline how humans and AI should divide work to get the best from both. 1) From Data to Wisdom: Where AI Sits in the DIKW Pyramid A helpful lens for thinking about AI’s role is the DIKW pyramid: Data : raw facts without context (e.g., “10, 6, 42, 8”). Information...

Why Most People Never Take the Leap “Two people. The same opportunity. One jumps. The other freezes. The difference? It’s in their brain chemistry.” Two people stand at the edge of the same cliff. Below is not a drop to certain death, but an opportunity—messy, uncertain, and full of potential reward. One jumps. The other takes a step back. We tend to call the first person brave and the second cautious . But neuroscience has a more precise explanation—it’s in the way their brains are wired. When we look at the landscape of business owners and entrepreneurs, one thing becomes clear: genuine risk-takers are rare. While many people dream of starting a business, only a small fraction act on it—and an even smaller group sustains it long-term. One explanation may lie not just in economics or social conditions, but deep inside our biology—specifically, how our brains are wired. The Brain’s Risk–Caution Spectrum Research in neuroscience suggests that ou...

It’s been nearly three years since ChatGPT’s sudden rise triggered both excitement and anxiety across the globe. Early reactions ranged from fears of mass job loss to visions of a fully automated future. Today, AI tools are part of everyday life — and while the initial panic has faded, we may be underestimating just how quickly things are moving. From Fear to Familiarity In a remarkably short time, people have grown accustomed to AI’s capabilities. We now generate images, write code, automate tasks, and even launch entire businesses with a few prompts — sometimes before breakfast. What once felt revolutionary has become routine. Forecasts and How They Went Wrong Last year, a study asked 67 experts when AI systems might surpass human specialists in niche fields. One benchmark: outperforming virologists on biothreat questions. The consensus forecast was 2034. Reality? It happened in early 2025 — a full decade ahead of schedule. Even “superforecasters” failed to...

For the past 50 years, computing has relied on one thing: the charge of electrons . From smartphones to AI supercomputers, everything we do digitally depends on pushing electrical charges through microscopic transistors. But what if we’ve only been using half the power nature gave us? Enter Spintronics —a transformational technology that could reshape how we compute by tapping into a second, overlooked property of electrons: spin . 🔁 From Charge to Spin: A Paradigm Shift Unlike charge (which flows), spin is a stable quantum property—an intrinsic orientation of electrons, like a built-in magnetic compass pointing “up” or “down.” Instead of physically moving charges (which takes time, energy, and produces heat), spintronics manipulates these spin states to store and process information in place. The result? Significantly lower energy consumption , faster data access, and much cooler devices —literally. 🧠 Memory and Logic in One Place — Like the Human Brain...

Recursion has always captivated programmers, engineers, and mathematicians with its elegant simplicity and powerful implications. By defining a problem in terms of smaller instances of itself, recursion can often simplify complex logic into a clear, concise solution. Nowhere is this principle more exciting—and potentially revolutionary—than in the evolving field of artificial intelligence (AI), specifically in the emerging phenomenon known as "recursive self-improvement." Artificial intelligence, despite its astonishing progress, seems relatively calm at the moment. However, beneath this apparent stillness brews an imminent storm. Many experts predict a coming "intelligence explosion," a point at which AI systems begin to improve themselves recursively and rapidly surpass human capabilities. Previously called the "singularity," this concept has been reframed as the intelligence explosion to avoid negative associations with cosmic singularities. ...

Agentic AI systems—large language models that can independently use tools—open extraordinary new possibilities, but they also introduce a novel security surface. Recent research shows they can (i) pick up “AI-worms,” self-replicating prompts hidden inside images or text, (ii) be hijacked by invisible instructions such as white-font email footers, (iii) discover serious software zero-days, and (iv) exhibit surprisingly self-protective or even coercive behavior in safety tests. Because data and instructions are blended in the same input, classic isolation techniques fall short; instead, layered defences, continuous red-teaming, and strict capability controls are urgently needed. 1. What “agentic AI” really means Agentic AI extends generative models by letting them plan goals and call external tools—browsers, email clients, code runners—on our behalf. IBM’s primer frames it as the next step after GenAI, shifting from “content creator” to “autonomous actor.” ibm.com Recent ...

Artificial-intelligence research, policy and investment have all accelerated in the first half of 2025. From the electricity that powers gigantic models to the ethics of synthetic media, the AI conversation now touches energy grids, labour markets, health care and search itself. Below is a concise roundup of the themes dominating headlines this month, with links to the most useful primary reporting. 1 | The power hunger: AI’s soaring energy bill Recent modelling shows that data-centre electricity devoted to large-scale AI could eclipse Bitcoin’s consumption well before the end of 2025, potentially topping 120 TWh a year.( techspot.com , mezha.media ) To stay ahead of demand, Meta has signed a long-term agreement with Constellation for nuclear-generated power to feed its Illinois AI facilities—one of several utility-scale deals now being struck by hyperscalers.( reuters.com , about.fb.com ) Chipmakers are also racing to curb watts per token. Nvidia’s new Blackwell GPU...

At Build 2025, Microsoft unveiled a bold vision for the future of the open, agentic web at scale. Moving away from vertically integrated application stacks, Microsoft is establishing a comprehensive platform designed to enable an open, scalable agentic web. This platform aims to unlock unprecedented opportunities for developers across every layer of the technology stack. Empowering Developers with Enhanced Tools Central to this transformation is Visual Studio , recognized as one of the most powerful IDEs for .NET and C++. Microsoft announced significant enhancements, including: Full support for .NET 10 Live preview capabilities during design time Enhanced Git tooling A new debugger optimized for cross-platform applications In parallel, Visual Studio Code marked its 100th release, introducing improved multi-window support and streamlined staging directly within the editor, reinforcing its status as a premier developer tool. Microsoft’s ecosystem further thr...

A ChatGPT deep-research paper. Entropy is a scientific concept most commonly associated with disorder, randomness, or uncertaintyen.wikipedia.org . From the steam engines of the 19th century to the smartphones and artificial intelligence of today, entropy and information theory have played pivotal roles in understanding how the world (and our computers) process information. In simple terms, entropy measures how unpredictable or mixed-up something is – whether we’re talking about molecules in a gas or bits in a computer file. This article will explore the origins of entropy in physics, how Claude Shannon borrowed the idea to create information theory, and what it all means for modern computing and AI. Along the way, we’ll use easy examples (like coins, messy rooms, and even a playful demon) to make these deep concepts accessible, and see how understanding entropy sheds light on the fundamental limits of computers . From Steam Engines to Bits: A Brief History of Entropy ...

Comparing an IT/software company to a football (soccer) club can be a helpful analogy for illustrating how different roles, strategies, and teamwork dynamics come together to achieve a shared goal. While no analogy is perfect, there are several parallels that can shed light on how these organizations function. Below are the key aspects and how they align: 1. Roles and Responsibilities Team Manager / Coach vs. Technical or Product Leads In a football club, the manager or coach is responsible for setting the team’s strategy, motivating players, and guiding their development. In a software company, technical leads or product managers fulfill a similar role by defining project goals, overseeing progress, and supporting their teams’ professional growth. Players vs. Developers/Designers/Engineers Each football player has a specialized position (goalkeeper, defenders, midfielders, forwards). Similarly, in an IT company, you have specialized roles—front-end developers, back-en...

The rise of China’s DeepSeek as a formidable competitor to OpenAI’s ChatGPT has sent shockwaves through the global tech industry. In January 2025, the startup’s AI assistant surged to the top of the Apple App Store’s free app charts in the U.S., China, and other regions, dethroning ChatGPT—a position the latter had held since its 2023 launch 12 7 . This milestone underscores DeepSeek’s rapid ascent, driven by cutting-edge innovation, cost efficiency, and a bold open-source strategy. 1. Benchmark Dominance and Technical Innovation DeepSeek’s flagship model, DeepSeek-R1 , has demonstrated performance rivaling OpenAI’s o1 model. On the international AI benchmark platform Arena, DeepSeek-R1 tied with OpenAI’s o1 in the StyleCtrl category (score: 1,357 vs. 1,352) and ranked third overall, surpassing models from Anthropic and xAI 1 12 . Key technical advancements include: Reinforcement Learning (RL) : Unlike conventional models reliant on supervised fine-tuning, DeepSeek-...