Learning to Code and Writing with Dyslexia

7

October

2025

No ratings yet.

I started using generative AI tools pretty late compared to most students I know. My first real experience with chat-GPT was when I began learning how to code. During that time I often ran into small logical errors, where I used wrong operators or mislabel a variable. Finding these small mistakes on my own can take a few seconds or sometimes 10 minutes, before you realised that you didn’t capitalize a letter in your variable. Chat-GPT was wonderful in not only finding these small errors but also in grading my beginner code and offering possible solutions to problems I faced, which made learning code a lot more dynamic, because of the instantaneous feedback that I could receive.

At first, I was hesitant to try AI for my writing, because I didn’t feel confident enough to distinguish good writing from bad writing. Having dyslexia makes spelling words and structuring sentences particularly challenging for me. Even when I understood a concept perfectly, I often struggled to convey it clearly in writing. Hyperbolically speaking I thank god everyday for spellcheck as I still have memories of writing my notes on pen and paper with spelling mistakes in every other sentence. Forming coherent structures was unfortunately still a problem for me and AI has been a real big help in conveying my thoughts to text. I put my uncoherent rambling into a prompt and the AI gives me a polished version. That I can use to as a guide to rewrite it on my own words. This process helps me convey the tone and meaning I had in my mind that I couldn’t express in my own words.

For me AI isn’t a “put in the prompt and I’m done” experience. Instead it acts like a guide to where I want to go. The output in most cases isn’t what I envisioned, but it helps me clarify my own ideas and get closer to what I intended. It provides a structure to arrive at the result I was aiming for.

Please rate this

Is the Stock market way overallocated on a gigantic AI bet?

19

September

2025

No ratings yet.

When watching the news, reading newspapers or watching podcasts, somehow there is always a article or topic about a company that spend a ridiculous amount of money on investments in AI. Whenever I come across one of these stories, I can’t help but wonder if such a large investment is worth it.

Take meta for example they reportedly offered Andrew Tulloch, the co-founder of Thinking Machine Lab, as much as 1.5 billion dollars over at least six years (Jin, B and Hagey, K, 2025) or the reported offer to Matt Deitke for 200 millions (The New York Times, 2025). All this money for just one person, how could this in any way be profitable?

But it is not just meta that are throwing large amounts of money in search of AI experts NVIDIA spent over 900 million to hire personnel from the AI company Enfabric (Kolodny et al., 2025) and Google reportedly offered 2.4 million billion dollars to Varun Mohan, the cofounder of Windsurf.

This strategy of spending hundreds of hundreds of millions dollars on the top guys of AI startups seems like betting everything on black in roulette, because it was your favourite colour. The thought that a few people, who had leadership positions in AI companies, are such geniuses that they are the key ingredient in revolutionizing the AI industry seems simplistic. In reality progress is not driven by the talent of the few, but it depends on workforces, infrastructure, and market readiness. The arms race of getting the top talent is only leading to inflating the costs without guarantying returns.

References:

Please rate this