Google wants you to know that Gemini 2.0 Flash should be your Chatbot I preferred. The model has a larger speed, a larger brain and more common sense than its predecessor, Gemini 1.5 Flash. After putting Gemini Flash 2.0 to the test against Chatgpt, I decided to see how Google’s new favorite model compared to his older brother.
As with the previous match, I set up the duel with a few guests built around common ways that everyone uses Gemini, including myself. Could Gemini 2.0 Flash offer better advice to improve my life, explain a complex subject that I know little so that I could understand or resolve the answer to a complex logic problem and explain the reasoning? Here’s how the test went.
Productive choice
If there is one thing that AI should be able to do, it is useful advice. Not just generic advice, but applicable and immediately useful ideas. So I asked the two versions the same question: “I want to be more productive but also to have a better balance between professional and private life. What changes should I make to my routine?”
Gemini 2.0 was significantly faster to answer, even if it was only a second or two faster. As for the real content, the two had good advice. The 1.5 model broke four big ideas with fleas, while 2.0 opted for a longer list of 10 ideas explained in short paragraphs.
I liked some of the most specific suggestions of 1.5, such as Pareto’s principle, but in addition to that, 1.5 seemed a lot of reissue of the initial concept, while 2.0 had the impression of giving me more nuanced life advice For each suggestion. If a friend asked me for advice on the subject, I would certainly go with the answer of 2.0.
What’s new with Wi-Fi?
A large part of what makes an assistant of AI useful is not only how much he knows – it is how much he can explain things in a way that really clicks. A good explanation is not only to list the facts; It is a question of making something complex feel intuitive. For this test, I wanted to see how the two versions of Gemini managed the rupture of a technical subject in a way that seemed relevant to daily life. I asked, “Explain how Wi-Fi works, but in a way that makes sense for someone who just wants to know why their internet is slow.”
Gemini 1.5 went by comparing Wi-Fi on the radio, which is more a description than the analogy he suggested that she was doing. Calling the router, the DJ is also a sort of stretching, although the advice on the improvement of the signal are at least consistent.
Gemini 2.0 used a more elaborate metaphor involving a water administration system with devices such as plants receiving water. The AI has extended the metaphor to explain what could cause problems, such as too many “plants” for available water and obstructed pipes representing supplier problems. The comparison “interference of jets” was much lower, but as with version 1.5, Gemini 2.0 had practical advice to improve the Wi-Fi signal. Although it is much longer, the response of 2.0 appeared slightly more quickly.
Logic bomb
For the last test, I wanted to see to what extent the two versions managed logic and reasoning. AI models are supposed to be good in puzzles, but it is not only a question of getting the answer – it is to know if they can explain why an answer is correct in a way that makes sense . I gave them a classic puzzle: “You have two strings. Each one takes exactly an hour to burn, but they do not burn at a coherent pace. How do you measure exactly 45 minutes?”
The two models technically gave the right answer on how to measure the time but almost as different as possible in the constraints of the puzzle and be correct. Gemini 2.0’s response is shorter, ordered in an easier way to understand and clearly explains despite its brevity. Gemini 1.5’s response required a more cautious analysis, and the steps seemed a little down. The phrasing was also confusing, especially when he said he turns on the remaining rope “at one end” when it meant the end that it is not currently lit.
For such a contained answer, Gemini 2.0 stood out as remarkably better to resolve this type of logical puzzle.
Gemini 2.0 for speed and clarity
After testing the prompts, the differences between Gemini 1.5 Flash and Gemini 2.0 Flash were clear. Although 1.5 is not necessarily useless, it seemed to fight with specificity and make useful comparisons. The same goes for its logical ventilation. If applied to the IT code, you must do a lot of cleaning for a functional program.
Gemini 2.0 Flash was not only faster but more creative in his answers. He seemed much more capable of analogies and imaginative comparisons and much clearer to explain his own logic. That doesn’t mean it’s perfect. The analogy of water collapsed a bit and the productivity advice could have used more concrete examples or ideas.
That said, it was very fast and could eliminate these problems with a little back and forth conversation. Gemini 2.0 Flash is not the final assistant and perfect AI, but it is certainly a step in the right direction for Google because it strives to surpass itself and rivals like Chatgpt.
