Wednesday, March 19, 2025

πŸ€– Wrapping Up Chapter 1: AI, Intelligence, and Philosophy

I’ve just completed Chapter 1 of the Helsinki University AI course, and it’s been an interesting mix of logic, debate, and deep dives into what AI actually is. This chapter wasn’t just about algorithms or machine learning—it was about philosophy, the definition of intelligence, and whether machines can actually "think."


πŸ”¬ The Turing Test and the Chinese Room

One of the key discussions was Alan Turing’s famous test, where if a machine can fool a human into believing it’s also human, it "passes" as intelligent. But does that mean it understands anything?

John Searle’s Chinese Room Argument challenges this. Imagine someone inside a locked room receiving Chinese messages and replying using a book of instructions without actually knowing Chinese. The responses might be perfect, but does the person inside the room understand Chinese? No.

This raises the question: If an AI can hold a conversation but doesn’t truly "understand" what it’s saying, is it really intelligent?


πŸ€” Strong AI vs. Weak AI

I also explored the difference between:

  • Narrow AI (Weak AI): What we have today—AI specialized in one task (e.g., ChatGPT, Google Maps).
  • General AI (AGI or Strong AI): The holy grail—an AI that can learn and adapt like a human.

At this stage, all AI is weak AI. Even self-driving cars don’t actually "understand" driving—they just process and react based on data.


πŸ“– The Definition of AI: My Thoughts

A big part of this chapter was defining AI, and while the given definitions were decent, I found them lacking. My take:

πŸ’‘ AI should be autonomous, adaptive, curious, and expansive.

  • Curious → It should explore ideas humans haven’t thought of.
  • Expansive → It should build on what it learns and grow.

πŸ“ Peer Reviews and Debates

One of the best parts of this section was reviewing and debating answers with peers. I reviewed 7 answers, and my own response received a 5.0 in its first review.

There was also the great AI vs. statistics debate in the previous section, where my instincts told me one thing, but ChatGPT led me down a different path. While I didn’t always get the "correct" answers, those discussions helped me learn why certain things are classified the way they are.


πŸš€ Moving Forward

With Chapter 1 complete, I’m excited to move into the next part of the course. The journey so far has been insightful, and AI is proving to be more than just code and algorithms—it’s about philosophy, ethics, and defining intelligence itself.

On to Chapter 2! 🎯

Tuesday, March 18, 2025

πŸš€ Second Section Down – Learning AI Through Debate

Today, I continued my AI journey with the University of Helsinki’s Elements of AI course. As always, ChatGPT remains my trusted study buddy, helping me debate, analyze, and challenge ideas rather than just memorizing facts.


πŸ“š Exploring AI's Related Fields

This section introduced key disciplines related to AI, including:
Machine Learning – Systems that improve with experience and data.
Deep Learning – A subset of ML focused on complex neural networks.
Data Science – A broad field incorporating AI, ML, statistics, and business applications.
Robotics – AI applied to real-world autonomous machines.

The first exercise tested my understanding of how these fields overlap. I had to correctly categorize AI, ML, DL, Data Science, and Computer Science in a Venn Diagram-style taxonomy exercise.

πŸ“Œ Final Score: 5/5! 🎯 Nailed it!


πŸ€– The AI Debate – When Logic and Course Answers Don’t Always Align

The second exercise required classifying AI applications into statistics, robotics, or machine learning. My initial choices were solid, but ChatGPT and I debated some of them—leading to some "incorrect" answers according to the course.

πŸ“Œ Final Score: 2/5. However, if I had stuck to my original choices, I would have scored 4/5.

Now, here’s the important part:
Even though ChatGPT’s suggestions lowered my score, I wouldn’t change a thing.

πŸ’‘ Why? Because the debate itself was where I learned the most.

Two key discussions stood out:
1️⃣ Gallup Polling (Statistics vs. ML) – My original answer was only statistics, but ChatGPT suggested adding machine learning. The course said ML was wrong, but based on real-world applications, I can still see how AI models could be used to analyze Gallup data beyond just statistics.

2️⃣ Missile Guidance (Robotics vs. Statistics) – I initially chose only robotics, but ChatGPT suggested adding statistics. The course marked it incorrect, saying guidance is mostly a physics problem, but ChatGPT’s reasoning made sense—there are statistical models used in trajectory prediction.

In both cases, the course followed a strict classification, but real-world applications are rarely so rigid. The discussions themselves helped me understand the nuances of AI far better than just memorizing the “correct” answers.


🎯 Key Takeaways from Today

Scored 5/5 in Exercise 2, proving a solid grasp of AI's related fields.
✅ AI is hard to define, but its key characteristics are autonomy and adaptivity.
✅ Machine learning is a subfield of AI, while deep learning is a subfield of ML.
✅ Robotics and AI often go hand in hand, but not all automated systems are true robotics.
Debating AI concepts is more valuable than simply memorizing answers.
Course exercises are useful, but real-world AI applications often blur the lines between categories.

Monday, March 17, 2025

Day One: Embarking on the AI Journey

 

Today marks the official start of my journey into AI. After thinking about it, planning, and figuring out how I want to approach this, I finally took the first concrete steps.


πŸ“Œ Registered for Helsinki University's Elements of AI

I signed up for the Elements of AI course by Helsinki University and MinnaLearn, [https://www.elementsofai.com/] a structured introduction to artificial intelligence. It offers ECTS credits, so I figured—why not? More importantly, it gives me a clear roadmap for understanding AI systematically rather than randomly consuming information.


πŸ€– My Study Buddy: ChatGPT

Since I don’t have a real-world study buddy (yet), I’m treating ChatGPT as my AI learning partner. It’s been an interesting experience so far, and we already had our first debate:


πŸ” AI or Not? The GPS Navigation Debate

One of the exercises asked whether a GPS navigation system that finds the fastest route is AI. My initial answer was no—after all, shortest-path algorithms like Dijkstra’s algorithm have existed long before AI. However, ChatGPT argued that modern GPS systems (like Google Maps) use real-time traffic data, machine learning, and predictive modeling, which introduce elements of AI. Since I am using Google Maps in my car, I thought that the correct answer for me would be "yes", and in the end, it turned out to be correct.  Truth be told, it was a tricky question, because if a GPS system does not take into account live traffic data feed, then the answer would be no, since finding the fastest route would involve just an algorithm, and not AI.


πŸ“– Quick Takeaways from Section 1 of Chapter 1: What is AI?

The first section of Chapter 1: What is AI? made one thing clear—AI is not easy to define. Many things we consider AI today may not be classified as AI in the future. However, two key concepts stand out:

  • Autonomy: AI can perform tasks without constant human guidance.
  • Adaptivity: AI can learn and improve over time based on experience.

These two characteristics help distinguish AI from simple automation.

At the end of the section, I completed an exercise where I had to classify different technologies as AI or not. I got 7 out of 7 correct, which was a great way to confirm that I understood the material.


πŸ’‘ What’s Next?

I have already asked ChatGPT to create a learning program for me—something I’ll go into in a future post. For now, my focus is on completing the first step of the learning program which is:  Elements of AI and making steady progress.

One section down—many more to go. πŸš€

Sunday, March 16, 2025

How I Decided to Learn AI in Every Shape and Form

A few weeks ago, I came across a statement from Mark Cuban, advising young people:

"Read books and learn how to use [artificial intelligence] in every way, shape, and form you can. It is a living library that gives you responses and can help no matter who you are or where you live."

The advice was meant for teenagers, but considering my plan to live for the next 100 years, I might as well treat myself as one. AI isn’t just another passing trend—it’s a technological revolution, and I don’t intend to sit on the sidelines.

I already have decades of experience with computers, from my early days with the ZX Spectrum, Atari CX, and 386 Turbo to working with SQL, COBOL, Pascal, and Oracle. I’ve seen technology evolve, from using punch cards and floppy disks to installing SUSE Linux in 2000 and recompiling the kernel just to get my Voodoo 3dfx graphics card working.

Yet, despite my background, I realized something critical: AI is an entirely different game.

I decided that if I wanted to truly understand it, I had to start from the ground up—treating myself as a complete beginner, no shortcuts.

Using AI to Research AI

Ironically, the first step in this journey was to use AI to plan my AI learning path. I relied on ChatGPT and other AI tools to:
Identify the best courses for beginners.
Break down AI concepts in a structured way.
Find certifications that would validate my learning.
Map out a long-term roadmap for mastering AI.

This quickly led me to another realization: I needed a way to document everything.

Why This Blog?

I didn’t want to just learn AI—I wanted to track my journey, document my mistakes, and ensure accountability. That’s how this blog was born.

This will be my personal guided Master’s in AI, completely self-directed, but structured.

The plan is simple:

  • Start from the basics.
  • Earn certifications step by step.
  • Apply what I learn.
  • Document everything.

Watching The Thinking Game only reinforced my decision. Seeing Demis Hassabis—who has Greek Cypriot roots—lead one of the most groundbreaking AI projects (AlphaFold, which earned him the Nobel Prize in Chemistry) proved that AI is a paradigm shift. This is as big as the discovery of fire or electricity.

Now, I’m fully committed to this path. I will learn AI in every shape and form, just as Cuban suggested. I will use AI to master AI.

And this blog? It’s where I’ll prove it.

Saturday, March 15, 2025

"The Thinking Game": A Catalyst for My AI Journey

At 20:20 today, my wife and I finished watching The Thinking Game on Prime Video. This documentary profoundly moved and inspired me, reinforcing my commitment to delve deeper into the world of artificial intelligence (AI).

The film chronicles the remarkable journey of Demis Hassabis and his team at DeepMind, highlighting their groundbreaking work in AI. Hassabis, born to a Greek Cypriot father from Famagusta and a Singaporean mother, grew up in North London. His heritage has been a point of pride, especially within the Greek and Cypriot communities.

In 2024, Hassabis and his colleague John Jumper were awarded the Nobel Prize in Chemistry for their development of AlphaFold, an AI system capable of predicting the 3D structures of proteins from their amino acid sequences. This breakthrough has been hailed as a monumental leap in scientific research, offering insights into biological processes and accelerating drug discovery.

Witnessing their dedication and the transformative potential of AI has solidified my resolve to embark on my own AI learning journey. The documentary not only showcases the technological advancements but also emphasizes the human spirit behind these innovations.

Kudos to Demis Hassabis and the entire DeepMind team for their contributions to science and for inspiring individuals like myself to explore the vast possibilities of AI.

For those interested in learning more about Demis Hassabis and his work, here's his Nobel Prize lecture:

https://youtu.be/YtPaZsasmNA

Friday, March 14, 2025

It's been quite a journey... Looking forward to the one just beginning!

I was born in the summer of 1969 (now 55 and semi-retired). My generation has witnessed some of the most radical technological shifts in human history—until the next one. We went from radio and black-and-white CRT TVs to experiencing the explosion of computers and the internet.

For us, knowledge came first from our parents and, if we were lucky, an encyclopedia at home. I still remember people coming to our house, trying to sell us books!


First Encounters with Computers & Gaming

My first experience with a "computer" was when one of my uncles returned from the United States, bringing a gaming console that connected to our black-and-white TV. It had only one game—PONG. I must have been around eight years old, and for me, it was absolutely mind-blowing! It was the most incredible thing I had ever seen.

My first gaming console was the Atari CX with cartridges. I was 11 or 12 years old.

By the time I was 15 or 16, I would take the bus downtown every Saturday just to visit computer stores and see the latest machines on display. I still remember the:

  • Oric
  • ZX Spectrum
  • Commodore 64
  • Texas Instruments TI-99/4A, which had an incredible game called Parsec and a speech synthesizer that could say "Good job, pilot!" in an almost human voice.

Programming & Early Computers

I got my first computer, a ZX Spectrum 16KB, but when I went to buy it, they didn’t have the 48KB version in stock. I had to wait months before I could finally upgrade it. That wait felt endless, but when I finally exchanged my 16KB to the 48KB version, it felt quite an upgrade since I was able to load much larger and better games that sometimes took about 30 to 40 minutes to load from the tape!

Programs were loaded using cassette tapes and a tape player. This was the machine where I learned my first programming language—BASIC.

College Days

Later, when I went to college, I got my first PC with an 8088 processor. I still remember debating whether I would ever need a 10MB hard drive. (Spoiler: I was wrong.)

On that computer, I had my first real gaming obsessionThe Ancient Art of War by BrΓΈderbund

By the time I was finishing college, I had upgraded to a 386 with a turbo mode of 25MHz.

This was the computer where I installed ORACLE and learned SQL, COBOL, Fortran, Pascal, dBase III Plus, and Clipper. It was also where I first worked with some of the most widely used software of the time, including WordPerfect (the predecessor to Microsoft Word), Lotus 1-2-3 (the predecessor to Excel), and the incredible Ventura Publisher

Ventura Publisher was the first desktop publishing software I ever used and my first experience with a graphical user interface (GUI). However, since mice were not standard at the time, it was designed to allow users to do everything using only the keyboard. For months, I worked with it without a mouse, handling every function entirely through the keyboard. Then, one day, I reluctantly decided to buy my first mouse and try it out. My productivity instantly quadrupled. It was unbelievable! From that moment on, I never looked back.

It’s also the computer where I installed my first Sound Blaster card, and for the first time ever, I heard sound coming out of my PC. The very first sound effect I remember? A creaking door opening from one of the Leisure Suit Larry adventure games. I had connected my Sound Blaster to my stereo, and the sound was unbelievable.

The game that really hooked me at the time was Dune by Westwood Studios. Probably the first awesome real time strategy game of all times. I played through all three factions (Atreides, Harkonnen, and Ordos) at least three times each. The soundtrack was absolutely amazing.  I found out recently that you can play this game online on a DOS simulator.

During my College Years I touched, and used:

  • Punch cards – The literal "coding sheets" that were fed into machines.
  • 8-inch and 5.25-inch floppy disks – Before the 3.5-inch floppies, these were the standard.
  • 3.5-inch floppy disks – My first Windows installation was Windows 3.1, which came on 11 3.5" disks.

By the time I finished my Bachelor’s in Computer Information Systems, I was already working professionally as a SQL programmer. That’s when I realized something important:

πŸ‘‰ I enjoyed programming, but I didn’t want to do it as a full-time job.


Army Days, University and the Internet beginning

At that time, there was no internet. Instead, we connected via 54kbps modems to BBSs (Bulletin Board Systems), where we would chat and play text-based games.

After completing my military service, I moved to the United States to attend George Washington University for my Master’s in Parallel and Distributed Computing.

By November 1995, as I was finishing my degree, the internet was just starting to emerge, but it was nothing like the internet we know today.

During my time in the US, I worked extensively with ADA and C++, and I even learned other obscure programming languages like SNOBOL. (which by the way I adored!)


Returning Home & The Internet Boom

In the summer of 2000, I installed my first SUSE Linux distribution, only to realize that my Voodoo 3dfx graphics card—brought back from the US—wasn't supported out of the box. This meant that I had to RTFM and manually recompile the kernel, a painstaking process filled with countless failed reboots. But when I finally saw the mouse pointer appear, signaling that the graphics card was properly recognized, I felt like I was the best programmer/hacker that ever walked the earth!

When I returned from the US, I also brought back a Palm Pilot, a cutting-edge device at the time. It had memory cartridges and a personal calendar appointment system—features that, back then, felt nothing short of revolutionary.

I first discovered the modern internet in the 2000s, long after my return home.

Right now, I am semi-retired, and I am looking forward to seeing what the next 10, 20, or even 30 years will bring.

I strongly believe that seven major technological fields will completely reshape our future:

  1. Biotechnology – Especially in extending the human lifespan.
  2. Quantum Computing – This will revolutionize everything and accelerate knowledge beyond belief.
  3. Driverless Cars – Automation will completely redefine transportation.
  4. Nanotechnology – Tiny machines, big changes.
  5. Renewable EnergyEnergy independence will reshape global economies.
  6. Artificial Intelligence – The reason I started this blog.
  7. Robotics – They will be the slaves of the 21st century, driving economy and one step closer to the the Universal Basic Income (UBI)

Why I’m Learning AI Now

I feel it’s time to get serious about AI.

I have been a subscriber to ChatGPT almost since its launch, and to be brutally honest, I am not impressed.

I expected way more considering the billions of dollars invested in AI research.

So I am embarking on a journey where, despite my long experience with computers, I will treat myself as a beginner and start from the ground up.

To stay consistent and accountable, I created this blog to:

  • Document my experiences
  • Track my knowledge and progress
  • Reflect on successes and failures
  • Maybe help others along the way

This is my Personal Guided Master’s in AI, and I’m excited to see where this road takes me.

🧠 Transfer of Consciousness: Moving Long-Running AI Projects Between Chats

When you work with conversational AI over time, you quickly discover an odd limitation: Your project stays in the old chat. The capabiliti...