Hey there,

Tired of the endless tabs and headaches when you’re looking for the best product to buy or need to find relevant sources for a paper? You’re definitely not alone — the traditional research methods can be exhausting. Luckily AI tools are here to make this process much smoother. 

Earlier this year, both Google and OpenAI released their own Deep Research modes. These AI research modes provide a deep dive into your queries and create a thorough, accurate, and well-crafted response from across the internet. Today we’ll be exploring what exactly Deep Research is and then we will compare the two most popular research tools: ChatGPT and Gemini. Honestly, Deep Research wasn’t something I’d used before, but now, they’ve become an absolute must for my more complex questions. 

What is Deep Research?

Forget asking AI simple questions or for some quick facts. New AI Deep Research tools are meant for bigger things. These unique AI experiences aren’t built for your average run-of-the-mill questions that you can easily Google (like where that one show is streaming this week); instead they’re built to provide in-depth exploration and analysis. Think of it as your very own AI research assistant who will tirelessly search the internet to find what you need.

Instead of pulling from their standard (and normally outdated) training data, Deep Research tools will actively go and search dozens (sometimes hundreds) of online sources to find the necessary information about your specific topic. They then combine all of this research into a detailed, well-crafted report that is grounded on the information it’s found.  

The goal is simple: save you time and sanity by going out and completing your research for you. This is actually an agentic feature that will self assess its sources to determine if there is any contradicting information and will automatically explore these further. I like to think of this like having a dedicated person who will read tons of content and then hand you their report. 

Accessing Deep Research & Understanding the Limits

Now that we know what it is, let’s see how to actually access it in ChatGPT and Gemini. 

ChatGPT

In ChatGPT, start by creating a new chat and clicking on “tools”. From here you can select “Run deep research” and you’re all set!

Gemini

In the Gemini app, you’ll again start by creating a new chat and then hit “Deep Research” at the bottom of the screen. 

Before you get hooked on using Deep Research in all your AI interactions, it’s important to understand that free users (of both ChatGPT and Gemini) will only get 5 Deep Research requests every month. Google has explained that this is due to the large amount of computational power needed to complete a single request. Users with Gemini Advanced will get 20 requests a month while ChatGPT Plus members get 25.

Deep Research In Action

Alright, now let’s see Deep Research in action. For this example, I'm using ChatGPT’s Deep Research. I asked for a detailed report on the most promising AI startups, a topic that requires sifting through a lot of information. 

The first thing ChatGPT did was ask clarifying questions. This is actually a much needed feature because most people will give overly general prompts like I did. These follow-up questions help the AI (and you) get a better sense of the scope of the research. 

Now we can dive into the report that was generated. It starts with a helpful table that provides a brief overview of each company, which is great for quickly scanning through the information.

After that, it offers a more detailed description of each startup and highlights the most important information (product, funding, etc.). 

The most crucial part of this report is the source list, which appears on the right side. Having each source be accessible makes it easy to click through to any sources to verify information or read it yourself. In this particular example, ChatGPT conducted 26 independent web searches and incorporated material from 36 sources when generating this report in only 5 minutes. 

Honestly, I was really impressed with the report I got. I’ve shared this example so you can check out the results for yourself. 

Practical Use Cases: 

Now that we know more about Deep Research, let’s explore how it can be used. Much like general chatbots, research mode can realistically be used for anything but I’ll try to highlight some of the most impactful examples.

• Academic Help: I wish I had this tool in my college days to help find and summarize research papers. Deep Research can be a great resource for finding background information when writing an essay. 

• Market Research: Deep Research enables us to have custom made reports that summarize entire markets and industry trends. 

• General Knowledge: Like most AI tools, Deep Research can be used for more casual or random tasks. Anything from understanding a complex technical concept to exploring a historical event. The reports you get will be more detailed, accurate, and thorough. 

• Shopping (Product Research): Whether you are looking to buy a new car or simply the best running shoes, knowing which product would be the best fit can be difficult. Deep Research can make this process easy by analyzing all your options and explaining the pros and cons in an easily digestible spreadsheet. 

• Personal Data: One of the most significant capabilities of Deep Research is that you can upload your own files and data. This opens up even more possibilities like having the AI double check your analysis and correct or expand on it.

Conclusion

So, what’s the takeaway from all this? Deep Research is a game changer for conducting in-depth research, whether you’re tackling something big like market research or simply want a list of the best skis. Deep Research’s ability to quickly find, analyze, and summarize tons of data from the internet makes me believe that the way we approach online research has been permanently changed. I think we will be seeing more tools like this in the near future. 

As I mentioned earlier, both ChatGPT and Gemini have their own versions of Deep Research (that they conveniently named the same thing). And while they generally share the same core purpose, there are definitely some significant differences between them. 

Ready to see how they stack up? Our next newsletter is going to be a direct comparison of ChatGPT’s Deep Research and Gemini’s Deep Research. We’ll explore their similarities, key differences, and I’ll give you my recommendation on which to use.