The rise of AI and automation has raised concerns about the potential loss of millions of jobs. In an interview with authors Daniel Abraham and Ty Franck, who wrote the science fiction series “The Expanse,” the potential for AI to replace jobs and the implications for a universal basic income (UBI) was discussed. Abraham and Franck explained that in their fictional world, Earth is jobless mainly due to AI and automation, and the government provides essential services to support the population.
While they acknowledged that their fictional “Basic” system differs from UBI, they explored the moral questions around assisting those who cannot find work. They also highlighted that the rapid development of AI and automation is pushing society closer to a future with too many people and not enough jobs, which may require additional support to keep people afloat. They noted that we are already seeing a growing economic disparity and are nearing a point where decisions will need to be made about how to support displaced workers.
NASA is implementing AI and citizen reporting to study UFOs in order to gain a better understanding of these phenomena. Interest in UFOs has increased in recent years, prompting NASA to address shortcomings in current data-collection practices. The agency is currently evaluating a report and its recommendations on the matter. In 2022, US defense officials released videos of unidentified aerial phenomena (UAP) during a congressional hearing. NASA and the Defense Department have stated that there is no credible evidence that these flying phenomena are of extraterrestrial origin.
However, the agency aims to determine whether these phenomena pose a potential threat to airspace safety. To support this investigation, Mark McInerney has been appointed to the newly created role of director of UAP research, where he will act as a liaison to the Pentagon and centralize communications and resources. NASA intends to use AI and machine learning, as well as crowdsourcing through public reporting, to increase the amount of data surrounding UFO sightings. The goal is to better understand already observed phenomena and identify any new ones.
Metaphysic, a generative AI tech firm, has announced its new system called Metaphysic Pro, which aims to help actors and individuals manage their personal data, including their likeness and voice. The system uses generative AI algorithms to allow actors to create their own face, voice, and performance data over time, and also helps them manage how it is used by third parties, such as consent, compensation, and copyright issues. Celebrities including Anne Hathaway, Octavia Spencer, Tom Hanks, Rita Wilson, Paris Hilton, and Maria Sharapova are said to be early users of the system.
This announcement comes at a time when the issue of how actors can control their personal information in the age of AI is a central concern, which has prompted the current SAG-AFTRA strike. Metaphysic aims to build a secure and transparent platform for artists, performers, and filmmakers to coordinate the use of personal data for AI content creation while ensuring consent and compensation. Metaphysic previously announced that it will be using its generative AI-driven Metaphysic Live tool to de-age actors, including Tom Hanks and Robin Wright, for an upcoming Miramax movie.
Google has reportedly granted access to a select group of companies to an early version of its conversational artificial intelligence software called Gemini. This move suggests that Google is nearing the integration of Gemini into its consumer services and plans to sell it to businesses through its cloud unit. Gemini is aimed at directly competing with OpenAI’s GPT-4 model, which has been generating substantial revenue for the startup. Financial institutions and other businesses are willing to pay for access to the GPT-4 model and the ChatGPT chatbot it powers.
By giving outside developers access to Gemini, Google is positioning itself to enter the growing market of conversational AI. This move will likely provide Google with a significant edge over its competitors. However, the exact timeframe for the rollout of Gemini remains undisclosed.
Google’s entry into this space indicates the increasing importance of AI technology for consumer services and business applications. The incorporation of conversational AI into Google’s offerings has the potential to significantly enhance user experiences and drive revenue growth for the company.
Microsoft has developed a tool called EvoDiff, which uses AI to generate proteins for drug design. EvoDiff is a general-purpose framework that can create “high-fidelity” proteins given a protein sequence. Unlike other protein-generating frameworks, EvoDiff doesn’t need structural information about the target protein. This reduces the computational and human resources needed to design proteins in the lab. EvoDiff could be used to create new enzymes for therapeutics, drug delivery, and industrial chemical reactions.
The framework is based on a 640-million parameter model trained on data from different species and functional classes of proteins. EvoDiff is a diffusion model that gradually subtracts noise from a protein sequence, stepping closer to a desired protein sequence. It can also fill in gaps in existing protein designs and synthesize proteins that do not fold into a three-dimensional structure. Although the research behind EvoDiff has not been peer-reviewed, the team plans to test the proteins generated by the model in the lab.
AI has made significant progress in accurately interpreting medical images such as x-rays, CT scans, and MRIs, as well as pathology slides and retinal photos. AI algorithms can detect features in medical images that humans cannot see, providing valuable insights into various aspects of human physiology and disease risks. Currently, AI models are primarily focused on image interpretation but have the potential to expand into text and speech analysis, creating multimodal AI systems. The development of multimodal AI requires self-supervised and unsupervised learning methods to handle the massive scale of input data.
One of the major advancements in AI architecture is the introduction of transformers, which allow models to process all data rather than relying on sequential feedback. Researchers have created multimodal AI models such as GPT-4, capable of working with text, audio, speech, and images by training on large-scale datasets from sources like Wikipedia and the Internet. Multimodal AI has the potential to revolutionize medicine by enabling virtual health assistants, remote monitoring, personalized treatments, and pandemic surveillance. However, there are still challenges to overcome, including biases, privacy concerns, regulatory approval, and the need for extensive evidence of benefit.
AI chatbots have achieved higher average scores than humans in a test designed to measure creativity, according to a study published in Nature Scientific Reports. Researchers asked three AI chatbots and 256 human participants to come up with creative uses for objects like a rope, a box, a pencil, and a candle within just 30 seconds.
The chatbots, including OpenAI’s ChatGPT and Copy.Ai, outperformed humans in terms of average scores, although the best human responses scored higher. The study raises questions about the unique characteristics of humans and the capability of AI to engage in creative thinking. Some experts argue that the AI models may not be demonstrating true creativity, but rather drawing on past training data. The findings highlight the need to examine the link between the tasks given to AI models and the cognitive capacity being measured. While AI has made significant progress in imitating human behavior, it is still not clear if it can replicate original thought.