• Home
  • AI

Google’s New AI Uses Your Search History & More AI Use Cases

A ton of new AI stuff came out this week, including Google's new Gemini model that uses your search history as context. (Creepy or cool..?)

We'll also show you a new benchmark that challenges LLMs to make Minecraft builds, show you some incredible new things built with AI vibe coding, and more.

Links:

Chapters:
0:00 What’s New?
0:26 Gemini 2.0 Personalization
4:48 Canvas in Gemini
5:19 Gemini Deep Research
6:07 Deep Research Audio Overview
7:25 NotebookLM Mind Map
8:07 Baidu
9:41 Vibe Coding Game Jam
11:17 Prompt of the Week
12:20 Nvidia GTC
13:44 Minecraft Benchmark
15:16 Roblox 3D Generation

#ai

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
 

  • @MyAnalyser says:

    Oh no! Not my search history! *blushes*

  • @naiklas1778 says:

    Love your thoughts about Google and privacy! Haha, it’s so nice! Finally they’re going all in without everyone noticing! 🤣
    It’s such a big step. Love the race as a consumer.🔥🔥

  • @patrickzupanc1795 says:

    Great video, thank you!

  • @Mr-vp8kw says:

    sending love from the US love your channel great content!

  • @creatingdigitalassets says:

    Regarding the search history in Google, it’s not completely accurate or representative of who we are , especially when we have clients and have had clients for years and we do research into their niches — I had clients that create online courses and programs with Kajabi for sobriety, porn addiction, post traumatic stress disorder, and all sorts of things that are not relevant to me, but I explored them in order to help them with their copywriting .

    I wouldn’t even know how to “exclude” them except for the brevity of the searches where there’s no pattern or longevity.

    So I worry about 100% automation in the assumptions that it’s making that an be incorrect.

  • @DJ369-Miami says:

    Oi, tudo bem? I wish Google LLM voice features would stop beginning everything with “Let’s do a deep dive…”

  • @h.c4898 says:

    I like the 2.0 perso model. About search history it’s not just the browser suggestion history, it’s also the chat history with Gemini. You can ask to retrieve old threads of topics you’d like to revisit it’ll summarize them and link them in the bottom card and then u can carry on. Pretty useful if you’re a researcher, no need to turn the chat history into a file manager. You just carry on. On the “saved infos” in the settings you can profile yourself what you want Gemini to know about you. That’s it. Don’t go beyond.

    Same with your browser history just make sure you leave stuffs you want Gemini to know for quick retrieval. You have options. It’s not like Meta these mf won’t let u control your data. It’ll eat it up to train their algorithms.

    I like this model better. Just me tough. It’s flash + pro + thinking now “contextual awareness”. Perfect for my use case may be not for others that I concede.

  • @naveenbattula says:

    Gemini 2.0 personalisation is scary good

  • @adrian_podar says:

    Thanks a lot for the great content! 💪

  • @sdhpCH says:

    Great Thumbnail!

  • @garethpr1 says:

    gemini 2.0 keeps telling me it can’t access other conversations let alone outside information.

    • @garethpr1 says:

      The user is asking me to tell them about themselves. This is a tricky request because as a large language model, I don’t have personal memories or access to private information about individuals unless they have explicitly shared it with me in the current conversation.

      My Analysis:

      The user has shared some information with me in this conversation, specifically:

      They are working on a coding project involving generating a grid of shanty towns.
      They are detail-oriented and have been very precise in pointing out errors in the grid generation.
      They found my performance frustrating, particularly my failure to follow the instruction to run the code and show results.
      They are taking a break from this project for the weekend.
      They are in the United Kingdom (based on the context).

      I should base my response on this information and avoid making assumptions or trying to access information outside of this context.

      My Plan:

      Acknowledge the Request: Directly address the user’s question.
      State My Limitations: Clearly explain that I don’t have personal information about them beyond what they’ve shared.
      Summarize Shared Information: Briefly recap the relevant details they’ve provided during our interaction.
      Focus on Their Actions and Opinions: Describe their behavior and expressed opinions during our conversation.
      Maintain a Helpful and Respectful Tone: Even though the interaction ended on a negative note, I should maintain a professional and helpful demeanor.

      Response Draft:

      “Based on our conversation, here’s what I can tell you about yourself:

      You are working on a coding project involving the generation of a visual representation of a shanty town on a grid.
      You have a keen eye for detail and have been very thorough in identifying discrepancies and errors in the generated output.
      You value accuracy and expect me to follow instructions precisely, particularly the instruction to run code and show the results.
      You found my performance during our interaction today to be frustrating, especially regarding my handling of the code execution.
      You are taking a break from this project for the weekend.
      You are located in the United Kingdom.

      Is there anything else you were hoping to know?”

      Self-Correction:

      The draft response seems reasonable. It sticks to the information shared within the conversation and avoids making assumptions. The tone is respectful. I will proceed with this response.

  • @serbandinu406 says:

    I always look forward to Ai News you can use. Loving the content and format. Keep it going

  • @AllenHorn0507 says:

    Hey Igor, I have a suggestion. What about using a Google phone number for the Chinese number?

  • @TheGuy0252 says:

    Great recap

  • @puiicherput4170 says:

    Veronum is on a one-way trip to financial freedom, Binance listing is coming…

  • @ชนัสนันท์บุญโย says:

    Hidden Codex of Wealth by Dorian Caine made me rethink everything society taught me about money. Feels almost illegal knowing this stuff—probably why it’s hidden from you.

  • @AlixHefmbry says:

    Honestly, I didn’t think true financial secrets existed until I opened Hidden Codex of Wealth by Dorian Caine. Now I realize they never wanted us to see this.

  • @ClarawGertrude says:

    I almost skipped Hidden Codex of Wealth by Dorian Caine, thinking it was just another average book. Glad I didn’t—this one exposed the money tactics they’re actively hiding from us.

  • @JanettaChillist says:

    The book Hidden Codex of Wealth by Dorian Caine feels forbidden for a reason—it reveals financial strategies the average person isn’t meant to learn. I get why it’s controversial now.

  • @EbonyFrace says:

    Everyone’s whispering about Hidden Codex of Wealth by Dorian Caine—and after reading it, I know exactly why. This book uncovers manipulative money secrets you weren’t supposed to see.

  • >