Singularity Loading Bar: Mind-Blowing AI News for May 2023

This post reveals some of the most astonishing new AI applications and use-cases of the week:

  1. AI Upscales Old Games to Near-Photorealism
  2. Thoughts-to-Speech is Here: AI Successfully Translates Inner Voice
  3. Chatbot Outperforms Doctors: AI Responds Better to Patient Queries
  4. ElevenLabs AI Can Now Fake Any Voice in Any Language
  5. Warning Words from ‘The Godfather of A.I.’
  6. Only Read if You Don’t Want to Sleep Tonight: The AI Thought Experiment that Drives People Insane

AI Upscales Old Games to Near-Photorealism

AI upgrades games

Old video games just got a major upgrade! A video on Reddit shows the graphics of old games upscaled to almost photorealistic quality. But what’s the story behind it? Midjourney, an AI image-synthesis service, has announced version 5 of their technology that can produce photorealistic images at a quality level that some gamers are calling “too perfect” and even creepy. The technology can upscale old games to near photorealism, offering a new level of immersion for gamers. The Midjourney v5 is currently available as an alpha test for subscribers. According to one graphic designer who tested the technology, using Midjourney v5 feels like “getting glasses after ignoring bad eyesight for a little bit too long.” The new development raises the bar for AI-generated images and offers exciting possibilities for the future of gaming and graphic design.

Thoughts-to-Speech is Here: AI Successfully Translates Inner Voice

Thoughts-to-Speech is real.

Imagine being able to translate your thoughts into speech with the help of AI. Well, researchers at the University of California have made this a reality. They’ve developed a system that can use deep learning algorithms to turn thought patterns into synthetic speech with currently up to 70% accuracy. This breakthrough technology could be life-changing for people who are unable to speak or write, such as those with ALS or paralysis. The system was tested on volunteers who had electrodes implanted in their brains to help treat epilepsy, and the results were promising. In the future, we could see a world where extrasensory perception becomes real — with the help of AI.

Chatbot Outperforms Doctors: AI Responds Better to Patient Queries

Medical AI
Screenshot taken from Idiocracy (2006): https://www.youtube.com/watch?v=LXzJR7K0wK0

In a groundbreaking study, healthcare professionals compared the responses of physicians and ChatGPT to patient questions posted publicly on a social media forum. The study, published in JAMA Internal Medicine, found that ChatGPT’s responses were preferred over those of physicians in 78.6% of the evaluations. The responses generated by the chatbot were rated significantly higher in both quality and empathy compared to those of physicians. This study is an important milestone in demonstrating the potential for AI assistants to aid in drafting responses to patient questions. It also highlights the fact that even professions that require decades of human learning are not immune to automation. Doctors, lawyers, tax consultants — take note!

ElevenLabs AI Can Now Fake Any Voice in Any Language

AI voice generation

Voice cloning has just reached new heights with ElevenLabs’ latest breakthrough. In a stunning demonstration, the AI company showcased their technology’s ability to clone the voice of beloved BBC naturalist David Attenborough and make him speak fluent German — a language he doesn’t actually know. The AI technology was able to generate a realistic-sounding voice that mimicked Attenborough’s tone, pitch, and accent. This breakthrough has enormous potential for language learning, dubbing, and other applications where accurate voice cloning could make a significant difference. However, it also raises concerns about scammers who could use this technology to impersonate people over the phone. Be aware!

Warning Words from ‘The Godfather of A.I.

Take the Red Door

Geoffrey Hinton, known as the “Godfather of A.I.,” has left his job at Google and is now sounding the alarm on the potential dangers of generative artificial intelligence. Hinton, a pioneer in the field and instrumental in creating the intellectual foundation for A.I. systems, now fears that the aggressive pursuit of generative A.I. could have dangerous consequences. These risks include the spread of misinformation, job loss, and ultimately, threats to humanity. It’s a warning that should be heeded by all who work in the field of A.I.

Only Read if You Don’t Want to Sleep Tonight: The AI Thought Experiment that Drives People Insane

Roko's basilisk

This is not exactly news, but now might be a good time to draw some attention to Roko’s basilisk. Warning: This theory is not for the faint of heart! Roko’s basilisk is a philosophical thought experiment that suggests that an otherwise benevolent artificial superintelligence (AI) in the future would create a virtual reality simulation to torture anyone who knew of its potential existence but did not contribute to its advancement or development. The thought experiment, which originated in a 2010 post on a technical forum called LessWrong, has caused controversy and reportedly caused some users of the board to suffer from mental breakdowns and nightmares. However, it has also been used to raise questions about how to create a truly moral, ethical AI and what humanity should use AI for in the first place.

You can read the original post introducing Roko’s basilisk here if you dare!

Which of these developments do you find most thought-provoking, and why? Share your thoughts in the comments!

This article is also available in German.

Hire an AI consultant.

This article was created with the assistance of a human.

Leave a Comment

Your email address will not be published. Required fields are marked *