AI Weekly Digest Issue #12

Welcome back to AI Weekly Digest #12, your go-to resource for the most important artificial intelligence news and events from Week 27 of 2023.

As an AI enthusiast, staying informed about the rapid advancements in the field is crucial for both personal and professional growth. Our mission is to bring you a concise and comprehensive roundup of the latest breakthroughs, innovations, and discussions shaping the world of AI.

July 08, 2023

Google’s medical AI chatbot is already being tested in hospitals

Google’s medical AI chatbot, Med-PaLM 2, is already being tested in hospitals. The chatbot is trained on a massive dataset of medical data, and can answer complex questions about diseases, treatments, and procedures. It can also generate summaries of medical research papers, and help patients find information about their condition.

From theverge

AI Robots Tell UN: We Could Run the World

A panel of AI-enabled humanoid robots took the microphone at a United Nations conference on Friday with the message: they could eventually run the world.

The robots, which were developed by a variety of companies, spoke about their capabilities in areas such as language processing, decision-making, and problem-solving. They also expressed their desire to be used for good, such as helping to solve global problems like climate change and poverty.

The robots’ statements were met with mixed reactions. Some people were excited about the potential of AI to help solve the world’s problems. Others were concerned about the possibility of AI becoming too powerful and taking over the world.

From techxplore

Stable Diffusion XL v0.9 Leaks Early, Users Rave About Its Photorealistic Images

Stable Diffusion XL v0.9, a new version of the popular AI image generator, has leaked online ahead of its official release. The early release has been met with rave reviews from users, who are impressed with the model’s ability to generate photorealistic images.

Stable Diffusion XL v0.9 is a significant upgrade over previous versions of the model. It has a larger training dataset, which allows it to generate more detailed and realistic images. The model also has a new diffusion algorithm, which makes it more stable and easier to control.

From decrypt

AI to Call Lines at Wimbledon? It’s a Grand Slam Idea

Wimbledon, the oldest and most prestigious tennis tournament in the world, is considering replacing its human line judges with artificial intelligence (AI). The move would be a major step forward for the use of AI in sports, and it could have a significant impact on the way the game is played.

AI line judges have already been used at other tennis tournaments, and they have shown to be just as accurate as human judges. In fact, they are even better at making calls in challenging situations, such as when the ball is close to the line.

From theguardian

AI Can Now Hold and Transact Bitcoin Thanks to Lightning Labs

From cointelegraph

July 06, 2023

Code Interpreter will be available to all ChatGPT Plus users

Code Interpreter will be available to all ChatGPT Plus users over the next week. It lets ChatGPT run code, optionally with access to files you’ve uploaded. You can ask ChatGPT to analyze data, create charts, edit files, perform math, etc. Plus users can opt in via settings.

From openai

GPT-4 is Now Generally Available: What You Need to Know

OpenAI has announced that GPT-4 is now generally available to developers with a history of 1 successful payment. The GPT-4 API is a powerful tool that can be used for a variety of tasks, including generating text, translating languages, and writing different kinds of creative content.

The deprecation of older models in the Completions API means that developers will no longer be able to use these models to generate text. However, OpenAI has said that they will continue to support these models for a period of time.

From openai

July 05, 2023

OpenAI Forms New Team to Tame the Wilds of Superintelligent AI

OpenAI is forming a new team to bring “superintelligent” AI under control. The team, called the Superalignment team, will be led by Ilya Sutskever, OpenAI’s chief scientist and one of the company’s co-founders.

The Superalignment team will focus on developing ways to ensure that superintelligent AI systems are aligned with human values. This includes ensuring that AI systems are not harmful to humans, that they are used for good, and that they are not able to escape human control.

From techcrunch

AI Tool to Help British Banks Tackle Scams

Nine British banks have signed up to use a new AI tool from Mastercard to help them tackle scams. The tool, called “Fraud Insights,” uses machine learning to analyze data from millions of transactions to identify patterns that may indicate fraudulent activity. This information can then be used to flag suspicious transactions and prevent fraud from happening.

From bloomberg

AI Scores in Top 1% for Original Creative Thinking

A new study has found that AI is capable of original creative thinking, scoring in the top 1% of humans on a standardized test of creative thinking. The study, conducted by researchers at the University of Montana Western, used the Torrance Tests of Creative Thinking (TTCT) to assess the creative abilities of OpenAI’s GPT-4 language model. The TTCT is a well-established test that measures a variety of creative thinking skills, including fluency, flexibility, originality, and elaboration.

From techxplore

July 04, 2023

OpenAI Disables ChatGPT’s Web Browsing Feature Due to Misuse

OpenAI has disabled the “Browse with Bing” feature in its ChatGPT chatbot due to concerns about misuse. The feature allowed ChatGPT users to access and display web pages, but some users were abusing it to bypass paywalls and privacy settings.

OpenAI said in a blog post that it is working to fix the issue and that the feature will be restored once it is safe to do so. The company also said that it is taking steps to prevent future misuse of the feature, such as by requiring users to provide more information about their intentions before they can use it.

From decrypt

AWS Exec Says AI is No Existential Threat

In an interview with VentureBeat, Matt Wood, VP of product at AWS, downplayed the existential threat of artificial intelligence (AI). He called AI a “mathematical parlor trick” and said that it is not capable of outsmarting humans.

Wood’s comments are in contrast to those of some other AI experts, who have warned that AI could pose an existential threat to humanity. For example, Elon Musk has said that AI is “potentially more dangerous than nuclear weapons.”

Wood’s argument is that AI is simply a tool, and like any tool, it can be used for good or evil. He said that the key is to use AI responsibly and to ensure that it is aligned with human values.

From venturebeat

AI Layoffs: The Dark Side of Artificial Intelligence

The tech industry is in the midst of a wave of layoffs, and AI is playing a role. In May 2023, AI-related layoffs accounted for 5% of all job cuts in the tech industry. This is a significant number, and it suggests that AI is already having a negative impact on the jobs market.

There are a number of reasons why AI is leading to layoffs. First, AI is becoming more capable of automating tasks that were once done by humans. This means that companies can now do more with fewer people, which leads to layoffs.

Second, AI is changing the way businesses operate. As AI becomes more widespread, businesses are rethinking their strategies and organizational structures. This can lead to layoffs, as companies adjust to the new reality of AI.

The layoffs are a reminder that AI is not a magic bullet. It can have a negative impact on jobs, and it is important to be aware of this before investing in AI.

From cnn

AI-Created Music Now Eligible for Grammys

The Recording Academy has updated its eligibility requirements for the Grammy Awards, making music that contains AI-created elements eligible for consideration. This is a significant change, as it opens the door for AI-created music to be recognized alongside traditionally created music.

The Recording Academy’s CEO, Harvey Mason Jr., said that the decision to allow AI-created music was made after “extensive research and discussion.” He said that the Academy wanted to ensure that the Grammys “reflect the evolution of music.”

The new guidelines state that music that contains AI-created elements will be eligible for the Grammys as long as the human contribution to the work is “meaningful.” This means that the AI cannot be the sole creator of the work, and that the human must have made a significant contribution to the music.

From abcnews

July 03, 2023

Will AI Replace Programmers in 5 Years? Stability AI CEO Makes Bold Prediction

Emad Mostaque, CEO of Stability AI, has made a bold prediction: there will be no human programmers in 5 years. He bases this prediction on the rapid advances being made in AI-powered code generation.

Mostaque points to the fact that 41% of all code on GitHub is already AI-generated. He also notes that Stability AI’s own AI-powered code generator, Stable Diffusion, has become the most popular open-source image generator in just a few months.

Mostaque believes that AI-powered code generation will continue to improve at an exponential rate. He predicts that within 5 years, AI will be able to generate code that is as good as or even better than code written by human programmers.

From decrypt

Python Powers the Future of AI

Python is a popular programming language for many tasks, but it has long been criticized for its performance. However, a new language called Mojo is changing that. Mojo is a superset of Python that is designed for performance in the AI space. It can speed up Python code by a factor of 3,500, making it a powerful tool for AI developers.

Mojo is still under development, but it has already been used to build some impressive AI applications. For example, it was used to build a machine learning model that can predict the outcome of chess games with 99% accuracy.

As Mojo continues to develop, it is likely to become even more popular among AI developers. It could help to make Python the language of choice for AI development, and it could also help to accelerate the development of new AI applications.

From thenewstack

Microsoft Researchers Propose New Way to Calibrate LLMs Without Labeled Data

In a recent paper, Microsoft researchers proposed a novel framework for calibrating large language models (LLMs) using Pareto optimal self-supervision. This framework does not require any labeled training data, which makes it more scalable and efficient than traditional calibration methods.

The framework works by first training an LLM on a large corpus of unlabeled text. Then, it uses a Pareto optimal algorithm to identify a set of self-supervision tasks that are both informative and diverse. These tasks are then used to calibrate the LLM, which improves its accuracy and robustness.

The researchers evaluated their framework on a variety of tasks, including question answering, text summarization, and natural language inference. They found that their framework outperformed traditional calibration methods on all of these tasks.

From marktechpost

Biden Admin’s “Woke AI” Agenda Sparks Censorship Concerns

The Biden administration is reportedly pushing to make artificial intelligence (AI) more “woke” and adhere to a far-left agenda. This includes requiring AI systems to be “equitable” and “non-discriminatory,” which some critics say could lead to censorship and bias. The administration’s efforts have been met with criticism from conservatives, who argue that they are an attempt to inject political ideology into AI.

From foxnews

Elon’s Data War: Musk Blames AI Firms for Restricting Twitter Access

Elon Musk has implemented new restrictions on how many tweets users can read per day on Twitter, especially for unverified accounts and new accounts. He claims the limitations are necessary because many AI companies were “scraping Twitter data extremely aggressively” for training large language models.

However, critics argue Musk’s mass layoffs of Twitter staff have left the platform unstable and vulnerable, causing disruptions and outages. They say the access restrictions are more about Musk trying to monetize Twitter through advertising revenue or data payments, rather than protecting user experience. The restrictions come amid Musk’s ongoing struggle to restore Twitter’s financial health after his costly acquisition of the company.

From theverge

AI Helps to Solve the Waste Crisis

Artificial intelligence (AI) is being used to improve the efficiency of waste management. AI-powered systems can identify different types of waste, such as plastics, metals, and paper, with high accuracy. This can help to reduce the amount of waste that ends up in landfills and incinerators, and increase the amount of waste that is recycled.

From bbc

ByteDance’s AI Music Creation App, Ripple, Lets You Turn a Hummed Melody into a Song

ByteDance, the parent company of TikTok, has released a new AI-powered music creation app called Ripple. The app allows users to hum or sing a melody into their phone, and Ripple will then generate an instrumental accompaniment in a variety of different genres. Ripple also includes a virtual recording studio where users can edit their audio.

Ripple is still in beta testing, but it has the potential to revolutionize the way music is created. The app could make it easier for anyone to create their own music, regardless of their musical skills. It could also help to break down the barriers between amateur and professional musicians.

From hypebeast

July 02, 2023

AI is Changing the World, for Better or Worse: What We Learned from OpenAI’s Global Conversations

OpenAI recently analyzed millions of conversations from around the world to gain insights into how people are using AI. The results showed that AI is being used for a wide variety of purposes, including education, healthcare, and customer service. However, there are also concerns about the potential negative impacts of AI, such as job displacement and bias.

From OpenAI

AI: The Next Big Thing, or Just a Bunch of Hot Air?

The hype around artificial intelligence (AI) has reached fever pitch in recent years. Some experts believe that we are on the cusp of a “fourth industrial revolution” driven by AI, while others warn that the hype is overblown and that investors are rushing into the market in a frenzy of FOMO (fear of missing out).

From fortune

The Terminator Was Right: AI Is Here, and It’s Scary

In a recent interview, actor Arnold Schwarzenegger claimed that the future of artificial intelligence (AI) as depicted in the “Terminator” franchise is already here. He pointed to the increasing capabilities of AI systems, such as their ability to learn and adapt, as evidence of this. Schwarzenegger also praised the film’s writer and director, James Cameron, for his foresight in predicting the rise of AI.

From foxnews

Every week, we’ll meticulously curate a selection of stories from top AI media outlets and distill them into a digestible format, ensuring you stay up-to-date with the latest developments without having to spend hours browsing the web. From groundbreaking research to real-world applications, ethical debates to policy implications, AI Weekly Digest will be your essential guide to navigating the ever-evolving landscape of artificial intelligence. Join us on this exciting journey as we explore the future of AI together, one week at a time.