Aidful News

Share this post

🤝 AidfulAI Newsletter #26: Exploring Visual Capabilities in AI

newsletter.aidfulai.com

Discover more from Aidful News

weekly ◆ AI news ◆ privacy-friendly AI ◆ PKM (Personal Knowledge Management) and AI
Continue reading
Sign in

🤝 AidfulAI Newsletter #26: Exploring Visual Capabilities in AI

Dr. Daniel Bender
Oct 19, 2023
Share this post

🤝 AidfulAI Newsletter #26: Exploring Visual Capabilities in AI

newsletter.aidfulai.com
Share

Dear curious minds,

Welcome to the ultimate newsletter for those interested in Artificial Intelligence (AI) and Personal Knowledge Management (PKM). In this week's issue, I bring to you the following topics:

  • ChatGPT: Image Input and Output, and Faster Than Ever

  • Google's SGE: Pioneering the Future of Search with Generative AI and Visuals

  • Revolutionizing Web Browsing: Arc Browser and Its Innovative AI Features

If nothing sparks your interest, feel free to move on, otherwise, let us dive in!

🖼️🏃 ChatGPT: Image Input and Output, and Faster Than Ever

  • Finally, the DALL-E 3 integration in ChatGPT appeared for me in the model selection window. I did not manage to sign-up for the Alpha. As of today, no option in the Beta feature settings visible, but the model is there and can be used. If you don't see this option yet, a re-login might help, but no guarantee.

    ChatGPT Plus subscribers should now have the DALL-E 3 model in the GPT-4 drop-down menu.
    I love the variation in the four images generated by DALL-E 3. On top, all hands look good, which a few months back was rare for generative models.
  • The option to upload an image and use it as additional input to your text prompt is also rolled out to all Plus users. To upload an image, click on the button on the left side of the input area.

    You can now use images as additional input to your text prompts in ChatGPT’s GPT-4 model.
  • An article states that the speed of GPT-4 is now nearly on par with GPT-3.5. The tests are done with the API from OpenAI. However, I could not reproduce this in ChatGPT where for me GPT-4 still is way slower, and not only because of the way GPT-3.5 did reply in my evaluation.

    GPT-3.5 found a shortcut for my speedtest as my instructions were contradictory.
  • My take: You know the saying, "A picture is worth a thousand words"? Well, it's like that – sometimes a picture can explain things better than many words can, and now ChatGPT can use them as input and even generate them as output for you. This is going to open up so many incredible possibilities.

🔍🎨 Google's SGE: Pioneering the Future of Search with Generative AI and Visuals

  • SGE, which stands for Search Generative Experience, is an initiative by Google aimed at enhancing the capabilities of its search engine with generative AI models.

  • So far, the SGE just used text models to generate answers to a user search.

  • In a recent update, SGE can now also generate images in addition to existing ones matching your search phrase.

    Google SGE shows generates images (right) besides images found with their search. [source]
  • The SGE is still in the experimental phase and so far, only available for US users opting into SGE in Search Labs.

  • My take: Google has many smart people creating new AI technologies. They also have many users, which is a big strength. By adding these new AI features to their search engine early, Google is making it tough for other companies to catch up. This move can make Google's search tool even more useful and hard to beat, keeping their users happy and staying ahead in the game.

🌐🚀 Revolutionizing Web Browsing: Arc Browser and Its Innovative AI Features

  • If you are a Mac user, you might come across the browser Arc from the browser company. It is a visually appealing alternative to the well-known browsers like Chrome or Firefox, which has a user-centered approach and offers various features aimed at improving the browsing experience.

  • There was a recent release of a bunch of AI features named Arc Max, which go beyond the default AI summarization features other browsers so far integrated.

    1. Ask on Page: Hold down Command + F on any page to ask a question and receive answers in seconds.

    2. 5 Second Previews: Hover and press shift over any link to generate a webpage preview without clicking.

    3. Ask ChatGPT: Type “ChatGPT” into the Command Bar, hit Tab, and get answers to questions in fewer clicks (requires a ChatGPT account).

    4. Tidy Tab Titles: Automatically renames tabs with tidier, shorter titles when pinned.

    5. Tidy Downloads: Organize files with smartly renamed downloads for easier retrieval later.

  • My take: I really want to try the Arc Browser, but so far, it is only available for macOS. However, a Windows version is announced to be released this winter. I am most excited about the Tidy Downloads feature, which would help me to clean up my download folder quicker. The Arc browser shows how generative AI can be integrated in clever ways to improve the user experience.

Thanks for reading! If you are not already subscribed, enter your mail address below to receive new issues and support my work.


Disclaimer: This newsletter is written with the aid of AI. I use AI as an assistant to generate and optimize the text. However, the amount of AI used varies depending on the topic and the content. I always curate and edit the text myself to ensure quality and accuracy. The opinions and views expressed in this newsletter are my own and do not necessarily reflect those of the sources or the AI models.

Share this post

🤝 AidfulAI Newsletter #26: Exploring Visual Capabilities in AI

newsletter.aidfulai.com
Share
Comments
Top
New
Community

No posts

Ready for more?

© 2023 Dr. Daniel Bender
Privacy ∙ Terms ∙ Collection notice
Start WritingGet the app
Substack is the home for great writing