•
July 13, 2023
•
6 mins
Hey Niuralogists!
This week we’re bringing you the newest updates that have been making waves within the world of AI and tech. Our main focus remains on exploring their implications for companies, businesses, policies, and individuals within the workplace. In this newsletter, we'll be focusing on the future retirement of OpenAI's older models, the temporary disabling of Chat GPT's internet browsing feature and more.
OpenAI has temporarily disabled the internet browsing feature of ChatGPT called “Browse with Bing”. This feature was designed to provide ChatGPT subscribers with up-to-date information by allowing ChatGPT to search the internet. However, it was disabled after reports stated it was generating information from websites protected by paywalls.
OpenAI is working to rectify the issues and plans to bring back the feature as soon as possible. This update comes as OpenAI is facing a multitude of legal challenges, such as a lawsuit in San Francisco accusing Open AI of violating privacy laws by secretly scraping 300 billion words from the internet, including personal information.
Humane, a startup founded by former Apple executives, Imran Chaudry and Bethany Bongiorno, has finally announced the name of its first product, “Ai Pin”, which is set to launch next year. The company, which was established in 2017, came out of stealth mode in 2021 and revealed the functionality of its device earlier this year.
The Ai Pin will be powered by Qualcomm's Snapdragon chips, which are commonly used in smartphones and mixed-reality devices. The device will primarily rely on voice-based control and will be able to understand instructions, act upon them, and deliver audio or visual responses. This functionality is made possible through a partnership with OpenAI, which has received investments from Sam Altman and Microsoft. Besides the fact that the device was powered from a small projector in Chaudry’s shirt pockets in the live demo, the final design of the device is yet to be revealed. Humane’s vision is to revolutionize the smartphone industry and create a new world powered by artificial intelligence and screenless technology.
In Japan, the education ministry recently released new guidelines that emphasize the need for students to understand artificial intelligence. These guidelines outline how generative AI can be integrated into schools and the precautionary measures needed in order to address these risks. The ministry encourages teachers to use AI in order to develop AI literacy and also streamline administrative tasks in order to reduce workload. The guidelines state that students should understand the advantages and disadvantages of AI, which include potential personal information leakages and copyright infringement. Additionally, it is stated that trying to pass off reports, essays, and other works produced by AI as one’s own is not allowed. The ministry is also working with the NHK to produce informative videos about AI to show to the students.
IBM and the All England Lawn Tennis Club have announced new AI-powered features for the Wimbledon digital fan experience. These features include generative AI commentary and AI draw analysis. The AI commentary feature, developed in collaboration with IBM iX, offers fans watching match highlights videos the option to receive audio commentary and captions of key moments.
This marks a step towards making commentary available for matches outside of Wimbledon’s Show Courts, which already have live human commentary. The AI draw analysis feature, on the other hand, provides a unique statistic to determine the favorability of the path to the final for each player in the singles draw. This feature leverages AI to consider factors such as match-ups against potential future opponents and the player’s position in the draw compared to competitors, providing tennis fans with deeper insights into the tournament
With the introduction of GPT-4, OpenAI recently detailed the depreciation of several of APIs that will retire by January of next year. From January 4, 2024, GPT-3 and its variations using the Completions API will retire to free up computing capacity. Applications using stable names for base GTP-3 models will automatically be replaced by their successors. Additionally, Developers using older embedding models will need to migrate to the new model released in December 2022, which is cheaper and more capable.
📬 Receive our amazing posts straight to your inbox. Get the latest news, company insights, and Niural updates.
AI may be changing the fitness industry thanks to the array of AI-powered fitness services such as Freeletics, Fitness AI, Aaptic, and Fitbod. These services offer personalized workout programs based on the user’s objectives and equipment and provide a cost-effective alternative to personal trainers. These apps use AI to analyze performance data, craft future workouts, and utilize data from other wearable fitness devices to provide a personalized view of the user’s progress. AI-powered fitness services are especially good for maintaining consistent workout routines and progressing, essentially mimicking a personal trainer. However, AI fitness tools have their limitations and fall short of providing the same level of interaction, support, and motivation as human trainers, and are probably not the best option for beginner lifters. Despite these shortcomings, these services show the way AI is improving and revolutionizing the fitness industry, leading toward a more personalized and accessible experience.
The idea of an AI becoming sentient became a hot topic when Blake Lemoine, a Google employee, suggested in 2022 that Google’s chatbot, LaMDA, had achieved sentience. However, experts still believe that LaMDA has not reached true sentience as it lacks originality and subjectivity. Other prominent figures in AI research, such as David Chalmer and Sam Bowman believe that we might witness sentient AI to some degree within the next couple of decades. While in theory, it’s possible to already digitize consciousness, the journey to sentient AI is still unknown. Despite these claims, the path to AI sentience also contains other hurdles including communication errors due to AI’s hard logic and possible loss of control over autonomous AI. Additionally, major concerns about AI gaining self-autonomy are that it may lead it to inquire about its rights, differences in AI versus human morality, and AI resisting being subjected to experiments. While the entire conversation regarding AIs becoming sentient remains speculative, conversations and research continue to grow.
A team of mechanical engineers, neuroscientists, and computer scientists developed a new and innovative way to calculate brain fluid flow using AI. The AI-powered measurements used in this project allow for a deeper understanding of the dynamics inside our brains. By integrating existing 2D data with physics-informed neural networks, the team has been able to create high-resolution insights into the perivascular spaces–a critical area linked to various neurological conditions such as Alzheimer’s. The new AI approach has the potential to expose pressures, forces, and the three-dimensional flow rate with higher accuracy than previous methods. While no advancements have been made in neurological treatment due to AI as of yet, this method shows a potential start to how AI can help.
.jpg)
