Future Tech Digest #7
Jan Tissler
CONTENTMEISTER ? – AI Content Strategy, AI Content Creation, Generative AI Workshops and Trainings, German Content, Translation and Transcreation.
Yesterday
Futuristic ideas and hypes from days past
While self-driving cars were once hyped to be ready by 2021, progress has been slower than expected. Many companies have shut down their self-driving projects in recent years. At the same time, Waymo and Cruise continue to make progress and are testing their technology in cities like Phoenix and San Francisco. Their technology has steadily improved, with Waymo's latest vehicles in Phoenix being smoother and more confident. While it will take time to make self-driving services profitable at scale, Waymo and Cruise plan major expansions in the coming years.
Today
What is happening right now
The European Union has passed a draft law that would be one of the first major attempts to regulate artificial intelligence. The law would restrict the use of facial recognition and require makers of AI systems to disclose more information. The draft law is seen as a potential model for other regions grappling with how to regulate AI. While regulating fast-moving technology is difficult, many argue that having no regulation at all would be worse. The draft law focuses on high-risk uses of AI and requires risk assessments before deploying certain AI systems. The final version of the law is expected later this year after negotiations between the EU's legislative branches.
Electric vehicle sales have been growing rapidly, doubling in the last two years. Research finds that this growth is due to improvements in EV features like longer range and lower costs, not an inherent preference for EVs. As battery ranges increase to over 300 miles and prices fall below gasoline vehicles, research predicts that a majority of consumers will choose EVs over equivalent gasoline models by 2030.
Computers that can simulate possible futures, even if just a few seconds ahead, represent a new paradigm of computing. Apple's new Vision Pro headset can place virtual objects in a room that stay in place even when the user leaves and returns, requiring the system to predict user actions. Ship sensors can detect incoming waves and choose optimal times to load cargo by simulating ocean waves ahead of time. Code autocomplete tools like GitHub Copilot suggest the next lines of code, allowing developers to choose from possible futures. As computers gain the ability to simulate the future in real time, it opens up possibilities for new applications that can warn of dangers ahead or optimize actions.
领英推荐
Tomorrow
The shape of things to come
The author had a demo of Apple's Vision Pro “spatial computer“ which allowed him to experience spatial photos, videos and virtual environments. He was impressed by the high image quality, eye and hand tracking, and ability to stay present in the real world while being immersed in virtual content. The author believes spatial video capture could allow people to relive precious memories in the future. While he has questions about Vision Pro's real world performance, he was excited about the potential for productivity and flexibility that it enables.
A solar power prototype launched into space in January has successfully demonstrated wireless power transmission in space for the first time. The experiment, called MAPLE, consists of an array of lightweight microwave transmitters that can focus energy on a desired location. MAPLE transmitted power to receivers within the experiment and also beamed detectable power to Earth. This demonstration shows that the flexible, low-cost technology required for a future space-based solar power system may be feasible. The ultimate goal of the project is to harvest solar power in space and transmit it to Earth to provide abundant, reliable energy.
Researchers at Nvidia created a Minecraft bot called Voyager that uses the advanced language model GPT-4 to solve problems and accomplish tasks within the game. Voyager generates objectives and code to improve its skills over time. Notably, the code generated by GPT-4 allows Voyager to learn how to make increasingly complex things and explore more areas of the game. The researchers argue that Voyager demonstrates the potential for language models to automate routine office tasks by figuring out how to perform actions on computers. While language models are often used for chatbots, using them to generate code and accomplish goals could have a bigger economic impact.
Editorial Team
Public Service Professional + Entrepreneur
1 年Winning editorial team!