DEEPSEEK ENTERS THE ARENA

IS IT USEFUL? IS IT SAFE?

- BY LINDA RUTH

When I first started experimenting with AI, only a couple of years ago, it felt like stepping into a world of magic. Back then, ChatGPT was impressive but limited—it couldn't browse the web or access real-time information. Now, it can pull the latest data, making it more dynamic than ever. But it's no longer the only game in town. We already pitted it against Google's AI model Gemini and found pros and cons for each. DeepSeek, a new contender, is making waves with its step-by-step reasoning, its transparent problem-solving—and its potential for privacy breaches.

WHAT MAKES DEEPSEEK DIFFERENT?

DeepSeek isn't just another chatbot—it's designed to think out loud. Its standout feature is its emphasis on chain-of-thought reasoning, which allows it to show its step-by-step thinking process. For example, if you ask it a complex question, it doesn't just spit out an answer; it walks you through its logic. This transparency makes it feel less like a black box and more like a collaborative problem solver. While this is particularly useful for math, coding, and scientific reasoning, it's also noticeable in everyday conversation.

This contrasts with ChatGPT, which also reasons through problems but tends to keep its process behind the curtain, though new settings offer more options. DeepSeek's approach has its pros and cons: if it makes an error, you can see where it went wrong, but the detailed explanations can bog you down. Take a question like: Why does my toddler make high-pitched noises? ChatGPT might (did) say: "Ah yes, the squeak phase—classic toddler behavior!" And then give a short paragraph on what could be the reason behind these squeaks.

DeepSeek might (did) say: "The high-pitched squeaking you're noticing in your one-year-old granddaughter is likely due to several developmental factors," and then provide a numbered list detailing each possibility, including when to worry and a conclusion paragraph.