AI Blog Summer Break Edition

Summer Break Edition

AI Blog

Issue #21

08/08/2023

Our Summer Recs

The August break is just around the corner, and we’re not about to let you hit the beach without an AI-themed reading list! Get ready to dive into mind-bending sci-fi classics, space operas with a twist, and thought-provoking reads that question the very nature of humanity.

Your next beach read:

Do Androids Dream of Electric Sheep? (Philip K. Dick, 1968): The inspiration for Blade Runner, this classic explores the line between human and machine.
The Moon is a Harsh Mistress (Robert A. Heinlein, 1966): A lunar colony rebels against Earth, aided by a superintelligent computer. Themes of freedom, revolution, and what it means to be a society.
Saturn’s Children (Charles Stross, 2008): A space opera with a provocative twist. Our protagonist? A sex-bot android left adrift after the extinction of humanity. Purpose, identity, and existence take center stage.
Ancillary Justice (Ann Leckie, 2013): This trilogy opener features a vengeful spaceship AI transplanted into a human body. Gender, colonialism, and revenge – oh my!
The Murderbot Diaries (Martha Wells, 2017): Meet Murderbot, the soap opera-loving cyborg with a conscience. Action-packed and surprisingly introspective, it paints a unique picture of humanity in a galaxy of danger.
Klara and the Sun (Kazuo Ishiguro, 2021): Klara, an Artificial Friend, observes her ailing human companion, Josie. A poignant exploration of love, connection, and what it means to be alive.
Power and Progress (Simon Johnson, Daron Acemoğlu): How technology has shaped prosperity throughout history, and the challenges societies have faced along the way.
Life 3.0 (Max Tegmark, 2017): Delve into the future of AI and how it could fundamentally reshape human existence.

AI Entertainment for the Voyage:

Westworld (2016-2022): Get ready for a wild ride through a futuristic theme park where the lines blur between human and host. Be warned: This one will have you questioning the ethics of AI exploitation in a whole new way.
Raised by Wolves (2020-2022): Imagine a mysterious planet where two androids are tasked with nurturing human children. Can AI truly provide the love and guidance needed for survival? Prepare for a thought-provoking journey.
I Am Mother (2019): Buckle up for this chilling tale. A lone robot, Mother, raises her daughter in isolation. But as secrets unravel, you’ll be left wondering about the true nature of AI and its potential for both nurturing…and control.
WALL-E (2008): Okay, okay, we know you’ve all seen this Pixar classic. But hey, what better way to remind yourself of the potential for AI to display genuine heart and resilience? WALL-E is a timeless reminder that even machines can dream of a better future.

Thanobots: The Digital Ghosts That Haunt Us

You’re spot-on about the ‘thanobot‘ concept. Think of character.ai chats, but fueled by the digital footprint of someone who’s passed away. It’s a chillingly logical step… but loaded with controversy. Why?

A thanobot might feel like a lifeline for the grieving – a simulated echo of their loved one. But the danger lies in blurring the line between AI mimicry and the real person. This can hinder the complex process of accepting loss, fueling denial instead of healing. Even the most sophisticated thanobot can’t replicate a person perfectly. The AI is trained on their online self, but everyone has layers. Discrepancies between the bot and the real person could lead to a whole new wave of pain and confusion for the user.

Did the deceased agree to have their digital essence turned into a bot? These ethical murky waters get even deeper when you think about how thanobots blur the line between human and machine.

Our desire to connect with the beyond is timeless – hence mediums, seances, and the like. Thanobots feel like the digital age’s answer, but there’s something unsettling about preying on vulnerability with technology.

Thanobots open a Pandora’s box of ethical and emotional dilemmas. Will they become tools for healing, or simply a high-tech way to deny the finality of death?

Perhaps is just us, but the productization of lamentation has always sat oddly with us.

Resources

AI Insider 📰

Meta plans to integrate AI chatbots into its platforms (Facebook, Instagram and Threads). The bots are aimed at increasing user retention and will have different personalities and skills. They will offer recommendations, help with search and general user entertainment.
- Meta also officially released AudioCraft as open source. It’s its all-in-one text-to-sound AI. The awesome thing is that all 20,000 hours of music used to train it are licensed!
- And Instagram plans to add labels to images that are AI generated or edited with AI.
OpenAI put out an official guide on how you can stop its web crawler from getting on your site and leveraging its content to train GPT-4.
Elon Musk bought the ai.com domain, which now takes you to xAI’s site. Meanwhile, not even six months had passed since OpenAI bought that domain.
YouTube is testing a new AI feature, which will make video summaries to help the user decide what to watch. These summaries will appear on watch and search pages and will not replace video descriptions.
Tinder is also testing AI features. Its own AI selects appropriate photos from the user’s photo album to help build their profile.
The need for data centers that can do compute-heavy AI applications continues to grow.
Google launched TextFX. It’s an experiment in collaboration with rapper Lupe Fiasco that intends to help artists with the creative process through the use of AI. It consists of 10 tools that offer features such as creating similes, breaking words into phrases, creating sensory details for scenes, etc.
LinkedIn is testing Microsoft’s AI art generator to design the visual elements of posts.
Zoom has updated its terms of service and plans to leverage user content to train its AI. A fact that caused outrage on Twitter – X, regardless of the fact that all AI companies are doing the same and stating it less overtly.

Learning Bytes 🧐

Midjourney is the #1 choice for creating AI-generated-images, but surprisingly we have not engaged with it as a business case study. In the video below, John Coogan does it for us.
Medium (blogging platform which we often honor here) requires all AI-generated content from now on to be flagged.
Similarly, the Publisher of Dungeons & Dragons announced that new guidelines will be set as they discovered AI-generated images in an upcoming book.
NASA and IBM created watsonx.ai, an open source AI model to advance climate science.
How easily can you tell if a text is written by AI? According to research from University College London, participants were able to identify AI text by 73%.