This Is The Best AI Tool Ever. I’m Not Kidding.

OpenAI just released ChatGPT Deep Research, a new tool that uses the full o3 model to complete complex tasks and answer difficult questions.

The benchmarks on release are seriously impressive, and our early testing supports the benchmarks. Watch the video for the full breakdown plus some good use cases and more!

Links:

Chapters:
0:00 The Best AI Tool
0:45 What is Deep Research?
1:21 How Does It Work?
3:09 Product Research
5:12 File Upload
6:51 Humanity’s Last Exam
8:34 Use Cases
9:46 My Work in Deep Research

#ai #chatgpt #openai

Free AI Resources:
🔑 Get My Free ChatGPT Templates:
🌟 Receive Tailored AI Prompts + Workflows:
👑 Explore Curated AI Tool Rankings:
🐦 Twitter:
📸 Instagram:

Premium Options:
🎓 Join the AI Advantage Courses + Community:
🛒 Discover Work Focused Presets in the Shop:

Joe Lilli
 

  • @cycologist8615 says:

    Some people spend time to save money, others spend money to save time. These tools are for the latter.

  • @japanimated9683 says:

    I hope Deepseek comes out with something so good chatgtp has to give it out free again.😂

    • @ai_handbook says:

      Definitely, deepseek gave that vibe to people 😅

    • @glomerol8300 says:

      Is Claude free? In any case, Ver. 4 (Feb 2?) seems overdue.

    • @itsjustme373 says:

      Now that open ai has released this tool, deepseek has something to copy. So theirs should be out in a month or two and it will have been created on an iPhone in less than a total of three hours.

    • @Lying29 says:

      The deepseek r1 as a free, offline-available model with full privacy of your data and query, is a better option to many users already.

    • @ji9306 says:

      Deepseek is far inferior tool guys. Let’s stop acting like the feature set is anything near the quality of OpenAI. I get it that everyone wants to root for the underdog, but the professionals comparing AI tools find a lot of limitations of Deepseek as it’s only a fork of Chatgpt’s 6 months old version AND O3-mini just launched with even cheaper API costs now, so that value prop is gone too smh

  • @johnupdate says:

    i have a question: let’s say i do a reseach regarding cameras and chat gpt searches the user reviews. can chat gpt recognize fake reviews as they are very common on amazon for example?

  • @icholakov says:

    open source alternative by the end of this week

    • @Ezkanohra says:

      Already exists.

    • @VraserX says:

      @@EzkanohraIt’s nowhere near as good as the OpenAI version. They simply lack the immense compute and power of full o3.

    • @icholakov says:

      @@VraserX so these difference is the underlying model not the process of compiling research reports. O3 has to be released for general use shortly or an open source alternative will take its thunder. The window of time between an innovation and old news is very narrow.

  • @llmtime2178 says:

    This thing is so good. I have good general knowledge about many topics but when I try to write substack-style posts on things I often have to do extensive research to try to find stuff Iremember reading but need the quote, but don’t remember where I read it etc etc. Using this I’mable to get a huge, densely-packed essay full of sources that I can freely draw on. It’s amazing.

  • @TheUberSchattenjager says:

    the top right of your camera’s sensor has two dust spots. 🙂 Also, really appreciate your guidance in this fast paced AI world

  • @Dr.Jekyll_ says:

    It saves hours for people that are really efficient at researching and reading, for most people it saves probably days.

    • @aiadvantage says:

      True. We discussed this yesterday and the tech savvy, 26 year old adhd types might be able to put together a big comparison table like that in 5 hours but for a 45 year old professional with decent computer skills that would take anywhere from 2 workdays to being an impossible task. Getting that for 2 $ is just crazy.

  • @oscarsalgar says:

    We need a comparison between this and Gemini’s deep research. I haven’t seen any yet. Everybody just assumes that this is much better, but I would like to know exactly why

    • @rodneyfletcher918 says:

      I’ve used both. Google’s version is good but it’s like giving a smart high schooler 24 hrs to research a topic and you get results back in 2-5 min. OpenAI’s version is like giving a smart post-doc 24 hrs to research your topic and getting results back in 5-10 minutes. Both are good but obviously one will be much better as far as breadth and depth of the knowledge..

    • @stoppls1709 says:

      gemini is useless garbage, feels like its gpt 3

  • @Drummagrl16 says:

    Since perplexity has been the leader in AI search (at least until recently), maybe you guys can do a video or perhaps talk a bit about how this deep research would compare to a perplexity pro search (maybe powered by one of the reasoning models)?

  • @dm204375 says:

    I think your videos are a fair assessment, and I would also like to see you cover the open source alternatives to all of the closed source counterparts. A lot of the time the open source stuff is close to the closed source quality and in some cases even better. This would bring a good balance to the discussion IMO. Because if you focus only on the closed source stuff eventually you start to look like a shill for a specific company or product. BTW I am not accusing you of anything, I think you are doing great just saying be careful not to fall in to that stereotype.

  • @prednosttrake says:

    Deepseek will come out with Deeper Research and query will cost you 2 cents.

  • @Tuzzz94 says:

    I can only imagine how shocked the academic world will be when they realize what this means (think about a student needing to write a research paper for a final exam and they just prompt o3, and the professor can’t tell)

    • @jason_v12345 says:

      Forget about the students. Conducting research is what PhDs themselves DO. It is what defines their profession.

    • @surajvkothari says:

      @@jason_v12345 Maybe OpenAI is essentially aiming to make PhDs in other fields redundant apart from those who are loyal to their company? Maybe eventually it will just be Sam Altman working alone at the company when they develop an AI so powerful it can just do everyone’s work at the company.

    • @adityagarg2185 says:

      @@jason_v12345 This is just synthesizing existing information though. Thats very valuable, but right now you still need humans to go out and collect new information. Thats exactly what PhD’s do.

  • @itsmeGeorgina says:

    Deepseek R1 is still best, the self learning capability is beyond remarkable

  • @2pacVivant0 says:

    i didn’t really get the difference between like asking GPT 4o or o1 to do an online research and summarize the results vs this ? like its just longer ? but who wants longer ? What’s wrong with the more concise answer we get on 4o ?

    • @surajvkothari says:

      For personal use, o1’s summary is fine. But if you work as an analyst in some company and your job is to do long reports, then this is what they require. Then again, if that is you, you’re made redundant, otherwise for personal fun questions o3-mini (free) is good enough.

  • @2pacVivant0 says:

    I think the revolution would be when GPT will be able to output real & usable PowerPoint Or Excel output file, with like proper templates and formatting on PPT similar to Gamma but more advance, or dynamic pivot table on excel etc

  • @lightlegion_ says:

    I love your vibe! Keep shining bright!

  • @jackstrawful says:

    ive been using Gemini’s Deep Research a bunch and been really impressed with it – i can scarcely imagine how much better, or qualitatively different, a Deep Research agent run on o3 must be compared with Gemini 1.5.❤

    Cant wait til it comes down to the $20 tier and i can give it all the same prompts for direct comparison

  • @ArrigoLombardi says:

    sorry for the random comment but y’all need to read The Censored Path to Wealth, seriously changes how you think

  • @lightlegion_ says:

    Hi, I’m Cecilia. It’s wonderful to meet you!

  • @TheoreticallyMedia says:

    Killer breakdown here, Igor!!

  • >