What Is AI?^[1]

Artificial intelligence (AI) used to be something people associated with science fiction movies, but today, it’s a part of daily life. Whether you’re using an AI-powered search assistant like Google Gemini, getting movie recommendations from Netflix’s AI algorithms, or even receiving AI-generated summaries of your emails, artificial intelligence is everywhere. If you have ever used a business’ customer support using a chat interface, the AI that powers the chatbot analyzes your question to provide an appropriate response.

But what exactly is AI? It’s just a fancy term for computer software that can do things that usually require human-like thinking. There’s a lot of different types of AI, but one of the most common is called machine learning. This type of AI helps us make sense of huge amounts of data by quickly identifying patterns and anomalies.

There’s also something called deep learning, which is a more advanced form of machine learning. This type of AI analyzes lots of information at the same time to make decisions. For example, self-driving cars use deep learning algorithms to understand the environment around them and make driving decisions.

The Rise of Generative AI

One of the biggest AI breakthroughs in recent years has been the rise of generative AI, which can create original text, images, code, and even videos based on recognizing patterns in vast amounts of data. ChatGPT, an AI-powered chatbot produced by OpenAI, was launched in November 2022. By January 2023, just two months later, ChatGPT was estimated to have reached 100 million monthly users. In comparison, Instagram took 2.5 years to attract 100 million users while TikTok took 9 months. The speed that users flocked to ChatGPT is a sign of how innovative, captivating, and powerful this AI technology is.

While some fear that AI will encourage academic dishonesty, many educators see it as a powerful tool for learning. Khan Academy’s AI tutor, Khanmigo, has been helping students by providing hints instead of direct answers. Khanmigo acts as a study buddy and tutor. It can do things like spot the mistake in an algebra problem and remind the student of the relevant concept that they forgot to apply. It can also have conversations with students about the significance of symbols in a literary work to help them better understand what they are reading. With this kind of one-on-one interaction, Khan Academy hopes to propel learning so that underperforming students will become high achievers.

AI Risks

As AI technology becomes more powerful, concerns about its misuse have grown. Early versions of ChatGPT could be tricked into providing instructions for harmful activities such as how to buy illegal firearms and make dangerous compounds with household chemicals. When alerted, OpenAI made changes so that ChatGPT no longer does these things. However, it’s impossible to eliminate all possible misuses, some of which could be extremely dangerous.

Advanced AI tools can also create hyper-realistic fake content, called deepfakes, making it increasingly difficult to distinguish truth from fiction. Malicious individuals or organizations can use AI to fabricate speeches, alter historical footage, or impersonate public figures, leading to potential political manipulation and reputational harm. AI-powered scams, such as deepfake voice fraud and chatbot-driven phishing schemes, are also on the rise, tricking individuals and businesses into revealing sensitive information. In response, governments and tech companies are working on AI detection tools, digital watermarking, and regulatory frameworks to mitigate these threats. This is why many believe that we should slow the release of such powerful AI tools until we better understand how they work, how they will be used, and what the consequences will be for misuse.

---