▾Artificial intelligence writes poems but struggles with math. Why can't ChatGPT and other chatbots handle even basic arithmetic? We reveal the causes of AI's mathematical mistakes, from tokenization which breaks numbers into unintelligible fragments, to the statistical learning approach that fails in mathematics.
Artificial intelligence, including ChatGPT, can write poems, compose music, and translate texts. Yet, it often stumbles on simple mathematical tasks. Why can't a chatbot, that handles complex language tasks, deal with math at an elementary school level?
Tokenization: When numbers break into pieces
One of the key problems is tokenization. This process divides data into smaller parts, called tokens. Imagine it like assembling a puzzle, where words are broken down into syllables. The tokenizer, the AI model responsible for this process, does not understand the meaning of numbers.
It may happen that the number 380 is perceived as one token, while 381 is perceived as two (38 and 1). This disrupts the relationships between digits and complicates the calculation.
A statistical machine that falters with numbers
Another reason for ChatGPT's mathematical difficulties is its statistical nature. The chatbot learns based on a vast amount of examples and looks for patterns in them. For instance, it learns that the phrase "Dear Sir" is often followed by the phrase "we are reaching out to you".
However, this approach faces challenges in mathematics. ChatGPT can guess that the product of numbers ending in 2 will end in 4, but it cannot handle intermediate results. Simply put, the ChatGPT model tries to guess the result based on learned patterns instead of performing a precise calculation.
The challenge of multiplication
A study conducted by Yuntian Deng from the University of Waterloo showed that ChatGPT struggles with multiplying numbers greater than four digits. The reason is that any error in a calculation step shows up in the final result.
Imagine it as a domino effect – one error triggers a chain reaction, and the result is completely off. However, there is hope that ChatGPT will improve in the future. Deng and his colleagues also tested the o1 model from OpenAI, which is characterized by logical reasoning capabilities.
This model achieved significantly better results than the standard GPT-4o and was able to correctly solve multiplications of nine-digit numbers. The o1 model thinks through the problem step by step, allowing for more accurate results.
Phone instead of a card or key. What is NFC and how does it work?
Just hold your phone up to a terminal and within moments the payment is made. However, NFC technology is capable of much more. We will explain what NFC is and how it works, what actually happens during NFC mobile payments, what else you can use it for, and how secure it is. We will also show what NFC tags are and how they can simplify some daily tasks.
What would happen if the internet went down for a day?
What if there was a full-day internet outage? An immediate collapse of society wouldn't occur, but daily operations would quickly become complicated because today's world is more dependent on the internet than we often realize. Which services would stop functioning first, and what would happen with payments, transportation, or work?
WPA, WPA2, WPA3: Which Wi-Fi security should you use today?
WPA, WPA2, WPA3, AES or TKIP? Wi-Fi security settings might initially seem unnecessarily complicated. However, when you understand what each abbreviation stands for, it becomes fairly simple. Therefore, we'll explain the differences between them and which is the best choice today. We'll also suggest what settings to choose if you have older devices at home that don't support the latest standard.
Is incognito mode really anonymous? Who still sees what you're doing
Do you think incognito mode hides you from everyone on the internet? In reality, it works a bit differently. We explain what it actually does, when it's sensible to use it, who can still see your activity, and whether history can be traced in incognito mode.
Why Does AI Hallucinate? A Peek Under the Hood of Language Models
AI can answer almost any question in just a few seconds. However, sometimes it confidently states information that is not true. Why does this happen and what are so-called AI hallucinations? In this article, we'll explain how large language models work, why they sometimes create false answers and how developers are gradually trying to mitigate this issue.
What is a CDN and why is internet content copied worldwide
Behind every loaded video, photo, or webpage is a technology called CDN. It is used by streaming services, social networks, and regular websites, yet many people have never heard of it. In this article, we will explain what this abbreviation means, how it works, why internet content is stored in various locations around the world, and why today's internet can hardly do without it.