never mediocre UI quality 🎨 again w/ AI

DEMO! and Kokoro, and another demo on deep Gemini Advanced research (agent)

Author

Martin Musiol
January 17, 2025 • Estimated Reading Time: 8 minutes

In partnership with

Hello all,

This time, I keep it short and sweet, as the newsletter suggests. Sorry for the occasional digressions.

  • Demo on how to perform deep internet research with Gemini Advanced

  • Kokoro - a groundbreaking text-to-speech model (small & performant)

  • DEMO: NEVER mediocre UI DESIGN again → 21st.dev + Cursor AI Agent 🔥

Let’s go!

Writer RAG tool: build production-ready RAG apps in minutes

  • Writer RAG Tool: build production-ready RAG apps in minutes with simple API calls.

  • Knowledge Graph integration for intelligent data retrieval and AI-powered interactions.

  • Streamlined full-stack platform eliminates complex setups for scalable, accurate AI workflows.

(✨ If you don’t want ads like these, Premium is the solution. It is like you are buying me a Starbucks Iced Honey Apple Almondmilk Flat White a month.)

[Demo] Deep Research with Gemini Advanced

Quick background story

I saw fellow AI enthusiasts use this awesome Gemini Advanced feature: perform deep research with Gemini (with Google Search, they have indexed the whole www anyways), and have their advanced Gemini 1.5 Pro model with a huge context window reason through it.

They shared well-written, generated reports synthesized from 1000s of webpages. Not only that, but with Gemini’s 1m token context window, and a reasoning workflow in the backend, it provides you an in-depth report after some minutes.

Me too? But I then realized that in Germany we don’t have access to it, and so made it public.

Gemini answered promptly (and publicly in the comments) and made me realize I had made a mistake.

You need to upgrade my Google Business Starter to Google Business Standard - 144$/year. I, like a good human, apologized. 🙇

How to deep research

First, upgrade to the Google Business Standard if necessary. Choose the Gemini Advanced - 1.5 Pro with Deep Research model, and enter your query. Here is my guideline video.

The screenshot captures a web browser window with Google Gemini’s interface prominently displayed. The interface is in “Gemini Advanced” mode, labeled “1.5 Pro with Deep Research.” At the center of the screen, there is a circular play button with a white triangle inside, indicating that a video or audio recording can be played. Below the button, text displays the playback speed as “1x” and the total duration of the recording as “4 min.” The timeline is set at the starting point, “0:00.” Sidebar Content: On the left-hand side, a vertical sidebar with a dark theme includes the following elements: 1. New Chat: A prominent button for initiating a new interaction. 2. Previous Topics: A list of previously saved topics, including titles such as: • “Challenges and Solutions for…” • “Marketing Tools: Boost Effi…” • “Quotes on Modern Blessings” • “Fehlermeldung: SPF-Eintrag…” • “Climate Action Commitment…” A “Show more” link below this section indicates additional saved topics. 3. Gems Section: Contains highlighted tools or functionalities, including: • “Chess champ” • “Brainstormer” • “Career guide” 4. Utility Options: Links for “Gem manager,” “Help,” “Activity,” and “Settings.” 5. Location Info: The user’s location is identified as “Munich-Neuhausen-Nymphenburg, Germany,” with an option to update it. Main Panel: The central section of the interface displays the Gemini logo and a placeholder for the playback feature. Beneath the play button, a text input field with a faint “+” icon is visible. The input contains the prompt: “Perform an in-depth research about how future-proof the German economy is, especially the automotive industry.” Browser Details: The browser shows several open tabs along the top, including services like: • Notion • Beehiv • Min Chat • Training… • HRS Gr… The user is likely multitasking across these applications. At the very top right of the screen, there is a profile icon and an orange notification indicator. The overall design is modern and clean, using a dark background with light-colored text and colorful accents to highlight interactive elements. The interface is focused on research and productivity tools, suggesting the user is engaging in professional or in-depth analytical tasks.

New Kokoro Text-to-Speech Model is Insane!

It has only 82M parameters, basically nothing.

This contributes to its super low latency. Meaning, how fast it responds.

It also outperforms large models, and is open-source!

Play with it on huggingface. Let me know if you implemented it!

Unlock the full potential of your workday with cutting-edge AI strategies and actionable insights, empowering you to achieve unparalleled excellence in the future of work. Download the free guide today!

(✨ If you don’t want ads like these, Premium is the solution. It is like you are buying me a Starbucks Iced Honey Apple Almondmilk Flat White a month.)

[DEMO] Never Mediocre UI Design Again → 21st.dev + Cursor AI Agent

Mamma Mia, how much time I wasted trying to optimize the layout and beauty of a product I built. That’s over now!

The Demo video is for Premium subscribers only. Or you bought the Everyone can Code!✨ AI builder course. Luckily, we have a deal.

Upgrade to Premium now (annual tier), and get the course for free!

I will never settle for mediocre design again. đź« 

I hope you enjoyed it.

Happy weekend!
Martin 🙇

Our webpage

You might like our last episodes:

Learn how to code with AI -> be unstoppable

Get Premium and receive AI course access on top! 🔥

mail.generativeai.net/p/learn-how-to-code-with-ai-be-unstoppable

first project doneâś… -> AGIpath(.net)

.. plus we will reach AGI this year

mail.generativeai.net/p/first-project-done-agipath-net

o3 and last tech updates of 2024

2025, big things will be built... also by you 🫵

mail.generativeai.net/p/o3-and-last-tech-updates-of-2024

Cursor 0.44 - Building Software Extremely Fast 🔥

.. for some too fast.

mail.generativeai.net/p/cursor-0-44-building-software-extremely-fast