(AI then vs AI now.)
the concept of AI is not a new one. like, not at all. Alexa and Siri have existed for decades. childhoods of playing Pac-man and Galaga have been thwarted by aliens and ghosts controlled by AI. IBM’s Deep Blue AI defeated chess grandmaster Garry Kasparov in 1997. in fact, the modern idea of “artificial intelligence” has existed over half a century, coined by computer scientist John McCarthy in 1956. so what is different about this AI?
(GPT has entered the chat.)
GPT, not to be confused with OpenAI’s product ChatGPT, stands for Generative Pre-trained Transformer. to oversimplify it, it’s autocomplete on steroids, except pre-trained on much, MUCH more data that is also aware of context (it knows the difference between a date on a calendar and a date in the produce aisle).
it predicts every word (technically, it predicts “tokens”, which can be words, parts of words, or punctuation, but “words” are more human.), one-by-one, based on the words that came before it. this is why most AI apps show you one word at a time. not only do you not know what it’s going to say next, it doesn’t know what it’s going to say next!
(don’t feel bad. the robots are completely winging it too.)
the first version (GPT-1) was created by OpenAI in 2018, but GPT-3 is what caught people’s attention in 2020. it was much more than auto-complete. it could answer questions (although it hallucinated a lot more than now.), summarize text, analyze sentiment, and it convincingly sounded so much more human than any other AI before it.
i personally remember when i used ChatGPT 3, i remember it feeling so unreal. the present started to feel like the future we’ve been promised in so many sci-fi movies. the reality of iRobot was imminent!
(the agentic age.)
agent (noun.): someone (or something.) that acts on behalf of another person or group.
an AI agent is exactly what it sounds like. instead of just answering the question “what is this weird mole on my back?”, a question you could have just as easily Googled, it will access your computer, launch a web browser, open a new tab, navigate to https://google.com, click the search box, type “what is this mole on my back”, click “Search”, go to the first WebMD link on the page, and call 911 on your phone because, apparently, you’re going to die.
(stupid stuff can now be automated!)
simply put, agents are AI with access to do stuff on your computer, apps, and other devices. you can ask it to do stuff for you, and it will do said stuff.
this is why Claude Code, a ChatGPT competitor, has become recently famous for. the tool is able to create slide decks, professional designs, and create websites + apps all by itself just by typing a few poorly written sentences.
this is when AI stops being a party trick and starts changing how society looks. many people’s jobs can now be automated. it gives ordinary people more power than ever before while threatening more livelihoods than ever before. it becomes a scary + exciting proposition
(what a time to be alive.)
so what’s next?
(every bubble pops.)
many experts are warning that the AI boom could pop soon because of the uncomfortable economics. companies are spending billions on data centers, chips, electricity, water, research, and infrastructure while offering millions of people access to powerful AI tools for as little as $20 a month.
the $20 you pay each month to ask ChatGPT or Claude, “can you rewrite this email so i sound professional but not the one to play with?” doesn’t come close to covering the full cost of building the models, expanding the data centers, developing the next generation of AI, and producing the enormous profits investors now expect.
it’s like a company giving (almost.) free rides in self-driving cars while spending billions to build the roads, manufacture the cars, and convince the entire world to stop driving for themselves.
that strategy can attract a LOT of customers very quickly. eventually, though, the company has to make the economics work. that could mean higher prices, stricter limits, cheaper models replacing expensive ones behind the scenes, more advertisements, or reserving the most powerful tools for businesses and wealthy customers.
the especially troubling part is that the AI industry is encouraging the mass population, including children, students, workers, and entire companies, to become increasingly dependent on tools whose current price and availability may not be permanent.
your $20 ChatGPT subscription is not necessarily losing money because of your usage personally, but the AI ecosystem surrounding it is being heavily subsidized by investors who expect to get their money back tenfold, and sometimes a hundredfold.
(and they want it back soon.)
(my thoughts.)
the times we are living through are interesting to say the least. we’ve been given a free taste of a drug and now the bill is coming due soon.
companies like OpenAI have limited options. they either go out of business (taking ChatGPT with them.) or they have to charge prices that most people cannot afford while giving less. what was a utility given to all of us could then just be a luxury that the ultra-rich could afford, further increasing the inequality where the rich get richer and the poor, poorer.
i don’t think it’s all doom + gloom, though. there are a lot of new competitors, especially in China, that have introduced comparable AI models at a fraction of the cost. just like most of our manufactured goods, China could become the world’s primary provider of ai.
another path is that many companies, like Google, have been working to make models small enough and capable enough to run on your computer or phone. Google’s Gemma 4 is already comparable to GPT 4.1 and it can run on a laptop (albeit a powerful one.) without hitting the servers of any AI company. i think this will continue to grow, especially for the privacy-minded because they don’t trust their most personal thoughts to a young VC-backed tech company whose only concern is making investors profit.
speaking of google, they also have a chance to win the AI race. unlike OpenAI (ChatGPT) + Anthropic (Claude) who are only alive because of investor dollars, Alphabet (Google’s parent company), is already a profitable $4.3 trillion company. it could afford to bleed money on AI for a while as it figures out how to make the economics work.
what happens next, no one can truly know, but what i can promise is that it will change the world forever.
how has AI changed your life. what was the moment when you realized the potential of what it could do? i’d love to hear your experience.
thanks for listening to my rambling.
(and to all the humans + robots, i love you.)
