Twelve years. That’s how long this blog has been running, a surprisingly long time, longer than many of my other endeavors. I can’t help but think about the fact that I started it on April 1st: April Fools’ Day, of all days. Maybe it’s a cosmic joke whose punchline still eludes me. Or maybe it’s just a conversation starter. Either way, the blog is here to stay.

This year, I made a bold promise: to write at least 100 posts. So far, I’m on track, publishing every other day since 2025 began. And somehow, it hasn’t felt overwhelming. In fact, it’s been fun.

Twelve years is also long enough to notice how much I’ve repeated myself. Many things that once seemed important no longer hold the same weight. The technology landscape has shifted countless times, yet here we are, still writing JavaScript. Back in 2013, AI wasn’t even mainstream.

After twelve years, I’ve started forgetting some of my older writings. While I stand by my past stances, my life has drifted in new directions, and with that, some convictions have softened. Technology, at its core, remains for humans. And hype always fades with a whimper.

Strangely enough, I’m enjoying reading my own blog. If nothing else, it’s a snapshot of my life. Even if this blog eventually fades away, it will always have at least one committed reader: me.

Happy 12th birthday, blog. Here’s to another twelve.

Everytime someone says "ok but don't judge me" I like to jokingly respond: I will absolutely judge you.

I never understood what it means not to judge. We absolutely judge people for everything.

• If you say something, you will be judged.
• If you say nothing, you will be judged
• If you show up, you will be judged.
• If you don't show up, you will be judged.
• If you write, you will be judged.
• If you don't write, you will be judged.

We judge books by their covers. We judge people by what they wear, how they talk, how they look.

The virtue is not in the lack of judgement. It's in how we react to judgement.

uBlock Origin Was Removed from Chrome... Not!

This morning, I was greeted with a message that uBlock Origin has been disabled on Google Chrome. It is "no longer supported." My very first thought was: "Over my cold dead body!"

uBlock Origin is my go-to extension. The same way I used to install Chrome for those pesky users still on Internet Explorer, I install uBlock on any browser I touch. I used to care that it breaks some websites, but not anymore. If uBlock breaks your website, it's probably for a good reason.

uBlock blocks ads. But why would I want to block ads in the first place?

Ads are invasive in more ways than one. They hijack the browsing experience, they track your every move, and they degrade performance. If you want to make a living from ads, that's fine, but count me out. If you want to block me from using your website because of this, that's fine too.

Google, being an ad company, has been threatening to get rid of uBlock for a long time. But they never said so directly. Instead, it's veiled behind "security measures," Manifest V3, and whatnot. The reality is, if ads weren't so bad, we wouldn't be using tools to block them in the first place.

My argument is simple: I will block ads. If you remove the tools I use to do so, as long as the computer belongs to me, I'll find other ways to block ads. It doesn't have to be rational—it's just what I'm going to do.

And no, it doesn't matter that Google doesn't "support" uBlock anymore. You can just re-enable it for now. If that fails in the future, you can always download it directly from the source and manually install it. If that fails, switch browsers.

We know what a web without ads looks like, and we are not going back.

Remote work has become the norm, and for many teams, seamless communication is essential to getting things done. But what happens when your primary tool goes down?

Just this morning, Slack was down, and I had a realization—my team had no backup way to communicate. No email thread, no alternative chat, nothing. We were completely cut off, waiting for Slack to come back online.

It wasn’t always like this. When I first started as a developer, communication wasn’t centralized. Everyone had their own preferred chat app. AIM, Yahoo Messenger, Facebook Chat, MSN, Pidgin. It didn’t matter. I used the Empathy client on Ubuntu, juggling Bonjour and Yahoo seamlessly. If one service went down, we’d just switch to another without skipping a beat.

But today? That flexibility is gone. Chat protocols are a solved problem, yet we now live in the era of walled gardens. Every service is proprietary, locking users into closed ecosystems. Just look at Apple, doing everything possible to keep iMessage from working with non-Apple devices.

Rather than just complaining about it, though, let this be a lesson: never put all your eggs in one basket. If your team relies entirely on a single communication tool, you’re setting yourself up for failure. Having a backup channel—even if it’s rarely used—ensures that when an outage happens, you’re not left scrambling.

So take a few minutes today to set up an alternative. Your future self will thank you when Slack (or Teams, or Discord) inevitably goes down again.

While U.S. tech giants are busy hyping AGI as some kind of all-seeing, world-domination-level superbrain that’ll make us all obsolete, Deepseek is out here flipping the script. They’re not building Skynet; they’re building something far cooler. AI for the rest of us.

Here’s how Deepseek puts it in their repo:

"Why? Because every line shared becomes collective momentum that accelerates the journey. Daily unlocks begin soon. No ivory towers—just pure garage-energy and community-driven innovation.

Maybe AGI is not what I thought it was. Maybe the "General" refers to the general public after all. AI for the general public.

The moment I heard about DeepSeek, I ran to check my Stock Portfolio to see how badly I was affected. Everyone and their grandmother are predicting the downfall of Western Civilization as we know it. Nothing has changed in my portfolio, because of course I was too late to invest in Nvidia. I am still late, but that's not the point here. The point is DeepSeek is the new Open Source model that rivaled ChatGPT on a shoestring budget.

It cost OpenAi upward of $75 million dollars to train their models. For each new iteration, this price continues to grow steadily. Those new models are proprietary to the company, and are not publicly available for anyone to use. We can refer to those models as closed source since only OpenAi has access to them. So when I heard DeepSeek is open source, and their latest model was trained on a budget of $6 million dollars, my first thought was to go ahead and download the training data and run it myself.

And that's where the open source term kinda falls apart. When we think of an AI model is open source, we swallow a gulp of air between open and source. We imagine access to the code, the data, and the inner workings. But in reality, the training data is never available. The data would include all the information scraped from the internet. Be it text, video, images, the majority of which is copyrighted. When we say a model is open source, we are referring to the weights generated through training. It is open weights. (Doesn't have that same ring to it, does it?)

For example, OpenAi or Deepseek will gather data around the web, build/train the model using this data, then call it AI. In the case of OpenAi, they keep the resulting model a secret only they can enjoy. Deepseek releases these models to the public as R1 and V3. Although they are a black box and we can't roll them back into the original data. However, we can still use these open models in our own devices. Anyone with the resources can download the model and run it.

It's not open source in the traditional sense where you have full access to the data. It's open source as in you have access to the model and its weights for free. We probably won't ever get an open source model in the real sense. But for now we will have to settle for "sorta kinda" open source.

As an aside, the market is responding to this news as if Nvidia has lost its edge now that we have a model that can be trained for cheaper. I think this is absurd. It looks more like we had an inefficient algorithm that required 100% CPU to run, now we've come up with a better algorithm that only takes 6% CPU. If anything, we can do more now with less hardware. In fact, now that we have all this hardware available, it will be trivial to do more training. I suspect we will see a leap in the LLama models in the coming days.

As the US moves to ban the TikTok app today, I couldn't help but think about a passage in the book Life of Pi.

In the book, Pi’s family is preparing to leave India, the only home they’ve ever known. The scene lingers with something profound: as the Patel family gets ready to board the ship that will take them to Canada, Pi’s mother, elegant in her finest sari, makes a small but significant gesture.

Mother was apparelled in her finest sari. Her long tress, artfully folded back and attached to the back of her head, was adorned with a garland of fresh jasmine flowers. She looked beautiful. And sad. For she was leaving India, India of the heat and monsoons, of rice fields and the Cauvery River, of coastlines and stone temples, of bullock carts and colourful trucks, of friends and known shopkeepers, of Nehru Street and Goubert Salai, of this and that, India so familiar to her and loved by her. While her men -I fancied myself one already, though I was only sixteen- were in a hurry to get going, were Winnipeggers at heart already, she lingered.

The day before our departure she pointed at a cigarette wallah and earnestly asked, "Should we get a pack or two?"

Father replied, "They have tobacco in Canada. And why do you want to buy cigarettes? We don't smoke."

Yes, they have tobacco in Canada-but do they have Gold Flake cigarettes? Do they have Arun ice cream? Are the bicycles Heroes? Are the televisions Onidas? Are the cars Ambassadors? Are the bookshops Higginbothams'? Such, I suspect, were the questions that swirled in Mother's mind as she contemplated buying cigarettes.

Today, for the first time ever, I've downloaded TikTok. I didn't bother creating an account. I just wanted to have it on my phone before it was removed from the App store.

Do I use TikTok? No. Will I use it if it is reinstated? Probably not. But I downloaded it anyway the same way Mrs Gita Patel wanted to buy cigarettes. It wasn’t about need or use. It was about the loss.

I've always advocated against tiktok. It is a time-sink that drowns our brains in a perpetual state of climax. Every video is designed to bring you to climax, and before it is done, the next video is loaded only to do the same.

But I've downloaded it because it was banned for the wrong reasons. TikTok wasn’t banned for ruining our brains or harvesting our data. It wasn't even banned because of National Security. No. It was banned because it didn't conform to manufactured consent.

The public's opinion may be completely irrational, but in a country that preambles with "We the People" we have silenced the voice of the many to please the few. If the ban is lifted tomorrow, it doesn't change anything. It's the very fact that it is possible to do so for "our protection". It makes you wonder: who’s really being protected here?

Sometimes, it’s not the app that’s dangerous. It’s what happens when we allow our freedoms to be erased, one ban at a time.

Blame culture is prevalent in many growing organizations and can lead to counterproductive outcomes. A developer introduces a bug, a product manager incorrectly describes a feature, or the marketer sends the wrong message.

In my experience, the organization always looks for a person to blame, and in extreme cases they get fired. But it’s almost always counter-productive to the company. For every mistake a person makes, there are tons of good work they do that goes unnoticed.

We are eventually all going to be replaced by robots. So let’s use an impersonal scenario where each person is replaced by an application.

Imagine an issue arises where an email has been sent to customers, and instead of a link to check the new service, an incorrect link is sent. To make matters worse, let’s say the link points to the competitor's service instead. To develop this feature, the research team surveyed customers. The VPs drafted a product spec, the product managers created a flow, the designers designed it, and the developers built it.

Somehow, the email was already sent to customers when the mistake was caught. It’s an embarrassment to the company. The company becomes a target of ridicule on reddit, for copying their competitor's homework and then linking to it. This makes the executive team extra mad.

When people are responsible for each task, we look for who messed up and fire them. Not just for the mistake, but for the embarrassment.

On the other hand, if robots perform, the blame will fall on the process. We will look at each point of the process and investigate what information was communicated incorrectly to cause the issue. We would add additional steps to verify the process and test for errors. If the issue was introduced between product managers and developers, we will look for ways to improve the communication. For example, we will add some tests between the two robots to make sure they each understand the request from the other.

It should be no different for people. When a mistake is made, the best thing to do is to review the process. Then find ways to improve communication between the two parties involved.

Ultimately, improving the process is key — don't hate the player; hate the game.

Or the perfect subscriber

Netflix can only get so many subscribers. Imagine for a minute that they get all members of a nation to subscribe. That means, they would have reached their maximum level of solvency. But because of the nature of market forces, at least in the US, Netflix has to continue to grow. They have to increase revenue every month, or at least show that they are working on a plan to increase it.

When you can't increase revenue, no one else to sell to, the obvious thing to decrease cost. Netflix does spend a boatload of making new shows. They will reach a limit, and their only option will be to decrease the cost of those shows. And hopefully maintain a similar quality. But you can only save so much by cutting costs before your customers start noticing. At the threshold, customers will stop paying you.

Another option will be to save on bandwidth, by decreasing the number of servers you have running. But a buffering video will lead customers to a more reliable service.

Ads can increase revenue, but customers that are already paying premium won't be happy. They are paying specifically to not see ads. Now you've given them a reason to leave.

These are three problems that a customer can face that will lead them to your settings page and hit that "Cancel Subscription" button.

The ideal Netflix customer should be one immune to these three problems. This customer does not care about the quality of your movie. High budget, low budget, no budget, it doesn't matter. At the end of the month, they pay for their subscription.

This customer does not care if when they click on a movie it instantly starts. If the movie buffers every 5 minutes, they don't care either. In fact, if they check the service and it says "Come back tomorrow," at the end of the month they will renew their subscription once more.

This customer does not care that the movie is now interrupted by ads. In fact, he doesn't even care that you increased the cost by 20%. They continue to pay consistently. This is the customer that will help Netflix continue to survive in a market that they have completely dominated.

This customer has simply forgotten that he has a Netflix subscription. He consumes no bandwidth because he doesn't even know he has the service. He doesn't mind the ads, because he doesn't know there are ads now. He doesn't care that you use cheap unconvincing VFX in your shows. He doesn't care that all your actors are C list. He doesn't even care that you are charging his card every month.

This is the perfect Netflix customer. In our algorithm driven times, a formula will be developed to make us consume less without ever stopping our subscription. Planet Fitness does the same, they are happier when you pay and don't show up. Manscaped does the same. And every subscription service will look for a way to get you to forget about them.

The screech of tires, the panicked cry of 'Donkey!', and the surreal slow-motion arc of a cart flipping through the air – that's the memory that greets me whenever I think of driving in Cairo Egypt. My first car accident, a baptism by fire in the controlled chaos of Egyptian roads, wasn't caused by a speeding taxi or a reckless motorbike. It was a donkey, and its cart, materializing out of my blind spot like a mirage. In the frantic search for merging traffic, my peripheral vision failed me, and my foot, guided by adrenaline, found the gas pedal instead of the brake.

The cart flipped, and I watched, horrified, as the donkey joined it in a brief, gravity-defying dance. The donkey, thankfully, survived its aerial adventure, though its loud, indignant bray still echoes in my mind. How, I still wonder, does a donkey end up on a freeway? And more importantly, how could a simple blind spot conceal an entire cart and its occupant? This wasn't just a driving mishap; it was a stark reminder of the limitations of human perception.

In Cairo’s orchestrated madness, where the unexpected is the norm, I realized how much we rely on our senses, and how easily they can betray us. It was a moment where the 'God View' of a self-driving car, that all-seeing, all-knowing awareness, would have transformed a near-catastrophe into a non-event. This isn't about replacing human drivers; it's about arming us with the tools to see what we can't, to anticipate the unexpected, and to navigate the unpredictable symphony of the road – whether it's on a Cairo freeway or a quiet suburban street.

AI drives better than humans, but here’s the thing: we don’t need self-driving cars—we need their tools. Imagine strapping the superpowers of a Tesla Autopilot or Waymo into your rusty sedan. Let’s talk about why drivers deserve the AI copilot’s cheat codes, especially if you’ve ever survived driving in a place like Cairo, Egypt.

Cairo: Where Chaos Meets Poetry (and Parking Angels)

Let me further paint the scene: Cairo’s streets are less “road” and more “organized anarchy.” Lanes? Optional. Traffic lights? Mild suggestions. Honking? The national language – a symphony of honks, ranging from impatient staccato bursts to drawn-out, almost mournful bellows. The thick, dusty air hangs heavy, layered with the scent of exhaust and street food. Motorbikes weave through impossibly tight spaces, a blur of motion in a kaleidoscope of faded paint on decades-old cars.

But here’s the magic: driving there feels like being part of a school of fish. No rigid rules, just flow. It's a ballet of intuition, where drivers anticipate each other's moves with a near-telepathic connection.

• Green light? A mere suggestion. Scan the horizon, wait for the collective nod, and then, and only then, dare to inch forward.
• Red light? If the traffic officer waves you forward like a conductor, go. Rules bend to the rhythm of the street.
• Parking? The moment you hesitate, a stranger materializes shouting “Erga! Erga!” (“Back up! Back up!”). They guide you into the spot, vanish, and leave you wondering if they’re a parking fairy. It's a system of on-demand assistance, a testament to the city's unique social fabric.

My mom never learned to parallel park in Egypt. She didn’t need to—Cairo’s hive mind sent helpers on demand. This "school of fish" dynamic, this unspoken communication, highlights the human element in driving – the intuition and adaptability that AI is still struggling to replicate.

Why We Need AI’s Toolkit (Not the AI Driver)

Self-driving cars get all the hype, but what if we stole their gadgets?

1. 360° “God View” Camera

Imagine seeing everything around your car—no blind spots, no guessing if that scooter is in your mirror. In Cairo, this would’ve saved me from merging into a donkey cart. Twice. This system blends data from cameras, radar, and lidar (think of it as a digital sixth sense) to map your surroundings in real time. It spots scooters, pedestrians, and yes, donkey carts, before your eyes even register them. This data is then processed and rendered, giving the driver an augmented view that extends far beyond the limitations of human vision.

2. Hazard Alerts for Humans

AI whispers: “Pedestrian detected, closing rapidly.” My ego protests, but my foot instinctively eases off the gas. AI, the silent guardian, preventing a near miss. These systems use predictive algorithms to anticipate potential dangers before they become apparent to the driver. By analyzing sensor data, they can identify patterns and predict the behavior of other vehicles, pedestrians, and obstacles. These predictions are then communicated to the driver through intuitive alerts, giving them valuable extra seconds to react.

3. The Parking Fairy, But Digital

No need to rely on mystical strangers. An AI overlay highlights the perfect parking angle and barks “Erga!” through your speakers. Mom-approved. This feature would use a combination of sensors and spatial mapping to identify available parking spaces and guide the driver through the optimal maneuver. Real-time feedback, both visual and auditory, would provide precise instructions, effectively digitizing the "Erga!" guidance of Cairo's parking angels.

AI Should Augment Humans, Not Replace Us

Self-driving cars are like overachieving interns: great at rules, terrible at chaos. But humans? We thrive in Cairo’s ballet of honks and hand signals. We just need better tools:

• Predictive honk detection (“Incoming road rage in 5 seconds!”). This system would analyze the acoustic properties of honks using machine learning. In Cairo, honks are a language. A friendly ‘beep-beep’ means ‘I exist!’ A prolonged ‘BEEEEEEP’ means ‘I’ve accepted my fate as a traffic ghost.

• Lane-finder radar for roads where lanes are conceptual. This technology would use high-resolution mapping and real-time sensor data to create a virtual, navigable path, even on roads where lane markings are faded, obscured, or non-existent. Because in some places, 'lane' is less a painted line, and more a philosophical debate.

• Traffic officer translator (“He’s waving you, not the bus!”). This feature could use image recognition and natural language processing to interpret the gestures and signals of traffic officers, providing drivers with clear, concise instructions, even in the most chaotic of intersections.

These tools wouldn't just be convenient; they'd be essential for navigating the complexities of modern driving, enhancing safety and reducing stress.

The Future of Driving? Think “Iron Man,” Not “Terminator”

AI doesn’t need to steal the wheel. Give us its X-ray vision, its spider-sense for potholes, and its encyclopedic knowledge of “is that a parking spot or a mirage?” Let humans handle the artistry—the intuition, the negotiation, the unspoken rules of places where GPS says “I give up.” We need a symbiotic relationship, where technology amplifies our abilities and compensates for our weaknesses, such as fatigue, distraction, and the inherent limitations of human perception. This collaboration would not only make driving safer but also more intuitive and enjoyable.

And if an AI ever replicates Cairo’s parking fairies? I’ll finally teach my mom to parallel park. Until then, I’ll keep dreaming of a world where my car shouts “Erga!” and a hologram gives me a thumbs up.

Let's demand these tools, and make driving safer and more intuitive for everyone.

TL;DR: Forget self-driving cars, give us their X-ray vision and spider-sense. Cairo proves humans rule the road, with a little 'Erga!' magic.