I tested Gemini’s new feature and it works just as I expected

Gemini 2.0 was released yesterday, and Google also announced an additional feature. It’s called “Deep Research,” and it’s technically part of the 1.5 Pro model, helping you complete research on a specific topic in minutes instead of the hours it usually takes.

This is a feature I’ve been waiting for because I have a lot of interests and in-depth research helps me learn about them faster. I tested it to see how well it worked and I have mixed feelings about it.

How to use in-depth research

Before we get into the hands-on part and my experience with the new deep dive tool, let’s talk about how to use it. First the bad news: Deep Research is part of the Gemini Advanced subscription, so you’ll need to pay $20 per month to use it. Google will eventually bring paid features to free users — just like it does with its products AI image generator – but the company hasn’t revealed any details so far. Additionally, Deep Research is currently only available online, but should be coming to a mobile app soon.

If you have a Gemini Advanced subscription, open the chatbot and select the “1.5 Pro with Deep Research” option from the drop-down menu at the top, as shown in the screenshot above.

Then the fun part begins. You can start with a very simple prompt, such as “Help me research the self-driving car trend for my thesis,” or you can be more precise with your request. After you enter your prompt, Gemini will build a plan that shows you the exact outline of the report it will create based on your research. You can continue to agree with it, or tell Gemini to change it, either by adding something or removing parts of it.

Then you need to wait a few minutes for Deep Research to finish its work. A full report will be provided upon completion of the study. If you need more information, you can ask follow-up questions. You can also open this report in Google Docs. As with all Gemini chats, it will be saved in your history for later use.

This is how in-depth research works on paper, but now it’s time to see how well it works in real life.

I want to start with something more complex. I am a financial junkie and love the stock market. I try to stay informed every day, but it’s not always easy. Some online resources provide me with a brief overview of what happened in the market the previous day, and if I have additional questions or want more in-depth data, I have to visit many websites to get everything I need.

I decided to try and see if deep research could give me all the data I wanted to see on one page, and then if I had any questions I could follow up with additional questions. Here are the tips I use:

Gemini I was quickly shown plans based on my prompts, also including more details such as exact dates – see screenshot above. After clicking “Start Research,” I had to wait four and a half minutes for the research to complete. This may seem like it takes a long time because we’re used to getting feedback from Gemini more or less immediately, but we have to remember that the AI ​​pulls data from 73 websites – imagine how long it would take you to do that on your own Own.

The report created looks great at first glance. It’s not very long and it doesn’t need to be as it should just be an overview. It includes several tables with sources under each paragraph, where you can check where Gemini pulled the material from. So far, so good. However, once I actually started reading the report in detail, the limitations of in-depth research became apparent.

But let’s start with the positives. The basic overview of the market as well as the percentage changes in the major indexes are very good. Aside from a few minor quibbles, the section on news and events affecting the market is solid, and the conclusion is well worth reading.

However, there are problems with almost every part. A table listing the performance of the top 10 S&P 500 companies by market capitalization does not actually include the top 10 companies by market capitalization. It shows only nine companies, and none of them rank among the ten largest companies in the United States. What’s more, it doesn’t show the percentage change for the four companies listed, it just says “not available.”

The accompanying text is even more confusing because it talks about year-to-date data rather than the data for the specific date I’m requesting, which is of no use to me. What I find particularly odd is this sentence: “An unknown company is leading the S&P 500 with a year-to-date performance of 308.05%.” This type of data in the stock market is usually transparent, so I don’t know why Gemini didn’t mention it to the name of the company. Google, what’s with all the secrets?

When looking at the sources, though, everything becomes clear. source Company names are not listed for free users because you must have an account to view them. Gemini didn’t take this into account and just said “an unknown company.” There’s a ton of other unimportant data in the various sections that feels out of place and doesn’t really provide much value.

However, I try to give Geminis second and third chances. I first asked it to study different styles of interior design and color schemes and which would best suit a small apartment in a modern building. This task is much easier for Gemini than stock-related tasks, and they do it really well. There were no silly mistakes this time, but I can say that the report provided is a bit basic and does not contain any images, which is important when talking about design. When asked for the image, Gemini said it was unable to provide it.

Then I asked it to research the topic Artificial Intelligence Chatbot Tell me which ones are out there, what they can and can’t do, how much they cost, and what their future is. Its results are similar to those of interior design. There are no major errors, but the research is a bit bland and reads like a standard post on the website. The report’s foreword even includes the following sentence: “This article delves into the world of artificial intelligence chatbots.” This seems to be something Gemini just pulled from the website it studied.

According to my experience extension of geminiI didn’t expect that the effect of Deep Research was exactly as advertised. I was right. It turns out that using it for stock market research on a specific day is generally more or less useless. While it did provide me with a lot of interesting data, it had enough bugs to convince me not to use it anymore. It did not provide me with the data I requested, and some of the insights included in the report were useless and should not have been included in the report.

When it comes to broader, simpler topics like AI chatbots and interior design, things are a little different. While the report provided is a bit bland and not very detailed, there aren’t any major errors in it, rendering the whole thing useless. You can always ask follow-up questions if you want more information, this part of the in-depth research works well.

Overall, in-depth research is a hit or miss thing. But we have to be realistic and take into account the fact that artificial intelligence is still in its early stages of development. The idea behind this Gemini tool is brilliant, and although it doesn’t work perfectly at the moment, it’s still useful enough in some situations. You just have to choose what it’s for.

But what excites me is the future of Google solving problems through deep research. Once improved, this tool will become invaluable in the lives of many people when they need to do in-depth research and don’t want to spend hours sorting through various sources.

What do you think of Gemini’s new in-depth research tool? Do you see yourself using it?