ChatGPT Plus at $20: A Habit Tax or a Real Necessity?

Let's be honest: the monthly twenty-dollar charge for ChatGPT Plus has become, for many of us, as automatic and unnoticed as the internet bill or a music subscription. We got used to paying for magic, or else we'd turn into a pumpkin under the limits of GPT-3.5. But the world has changed. While we were asleep, OpenAI turned the free version into a beast that can do almost everything the paid account does. A reasonable question follows: isn't it time to stop bankrolling Sam Altman's ambitions out of sheer habit? If the gap between free and paid access used to be the size of the Grand Canyon, today it is more like a barely visible crack in the asphalt.

Remember how things looked a year ago. Free users were stuck with a slow, slightly dim GPT-3.5 that got lost in three pines and produced C-student prose. The paid subscription gave access to GPT-4, the crown jewel that solved problems, wrote code and seemed almost intelligent. Today the situation is different. GPT-4o is available to everyone. Yes, free users have message limits, but for an ordinary person asking a chatbot to draft a workout plan or check an email in English, those limits are more than enough. You get the same speed and the same intelligence for absolutely nothing. OpenAI itself created a situation in which its flagship product started cannibalizing subscription sales.

So what are we still paying those two thousand rubles a month for? The main argument today is access to the o1 series of models. They can "think" before answering and solve very complex logic problems. But let's be objective: how often do you need to prove mathematical theorems or hunt for bugs in the architecture of distributed systems? For 90% of everyday tasks, o1-preview is overkill, and sometimes slow. If your work isn't tied to deep programming or scientific research, you are paying for a supercomputer and using it as a calculator. And that's without mentioning that the o1 limits still bite even on the paid plan.

We also can't forget the competitors. While OpenAI tries to keep us inside its ecosystem, Anthropic offers Claude 3.5 Sonnet, which many consider more "human" and smarter at writing. Claude has an excellent free tier too. If you use AI only occasionally, the combination of free ChatGPT and Claude will cover all your needs. You get the best of both worlds without spending a cent. In that reality, the Plus subscription starts to look like a tax on laziness: the laziness of switching between browser tabs.

Of course, there are people for whom Plus is vital. These are heavy users who generate dozens of images in DALL-E 3 every day, use custom GPTs to automate their business, or simply can't live without Advanced Voice Mode. For them, $20 is a laughably low price for a personal assistant that never sleeps.

But if you open the chat three times a day to rephrase a headline, you are simply handing money to one of the richest companies in the world. What's more, OpenAI recently introduced a Pro tier at $200 a month. That is a clear signal: the company is sorting users into "ordinary" and "professional".

And if you aren't ready to pay two hundred dollars, perhaps you should ask yourself whether you belong to the group that is perfectly well served by zero.

The market for AI subscriptions is oversaturated, and the value of each individual feature is falling. We are entering an era in which intelligence becomes a cheap commodity, almost a utility. Paying a premium price for it makes sense only if the tool directly increases your income or saves you hours every day. In all other cases it is just one more line on a bank statement that warms the soul with a sense of taking part in technological progress but delivers no real benefit.

Bottom line: Most individual users no longer need ChatGPT Plus. The free version of GPT-4o and competition from Claude have turned the $20 subscription into an excessive luxury. Are you ready to hit Cancel as an experiment?