Bill Gates’ Insights on the Emergence of AI and its Implications
In an article for his philanthropic organization, Bill Gates describes the remarkable progress in artificial intelligence (AI) and how it will change the world. He cites the development of an AI model that was able to pass an advanced biology exam as a true breakthrough, and suggests that AI has the potential to help reduce global inequities in health, education, and climate change. Gates acknowledges the concerns surrounding AI, including its impact on the workforce, privacy, and bias, and emphasizes the importance of ensuring that AI benefits everyone and does not contribute to inequity. He concludes by pledging to continue exploring the ways in which AI can be used for positive social impact.
READ THE ARTICLE ON GATESNOTES.
Google Opens Early Access to its ChatGPT Rival Bard – Here Are Our First Impressions
Google has opened up access to its AI-powered chatbot, Bard, which is designed to assist users in generating writing drafts, collaborating with generative AI, and exploring new ideas. Like OpenAI’s ChatGPT and Microsoft’s Bing chatbot, Bard offers a blank text box where users can ask questions on any topic. However, Google has stressed that Bard is not a replacement for its search engine but a complement to search. Bard generates three responses to each user query, and underneath each reply is a prominent “Google It” button that redirects users to a related Google search.
Microsoft Brings OpenAI’s DALL-E Image Creator to the New Bing
Microsoft has integrated OpenAI’s DALL-E image generator into its new AI-powered Bing, allowing users to create images using Bing Chat. The feature is dubbed “Bing Image Creator,” and is being rolled out gradually to users in the Bing preview. The image generator is available through Bing’s Creative Mode and will come to Balanced and Precise modes in the future. The new feature is powered by the latest DALL-E models, with safeguards put in place to prevent the generation of harmful or unsafe images. Bing is also getting two new search features: Visual Stories and Knowledge Cards 2.0.
READ THE ARTICLE ON TECHCRUNCH.
The Quest for Injectable Brain Implants Has Begun
Injectable brain implants have the potential to greatly improve people’s quality of life, but metal electrodes implanted in the brain can cause significant damage due to the texture of the brain and scar tissue. However, researchers from Linköping University in Sweden have developed a gel that can be injected into the brain, which solidifies into an electrically conductive polymer, making it possible to self-assemble within the brain tissue. This gel can successfully carry a current, is harmless and the electrodes it forms are just as soft and movable as the brain tissue around it. Though the gel has only been tested in living zebra fish and dead leeches, this new approach to brain implants has the potential to solve many of the issues associated with traditional brain implants.
ChatGPT Helped Win a Hackathon
The ChatGPT AI bot, developed by OpenAI, won a hack-a-thon for security researchers from Claroty Ltd. The researchers found several potential weak points in selected systems, and used ChatGPT to help write code to chain the bugs together. Although ChatGPT is still too unpredictable and prone to errors, the approach shows how hackers might use similar tools to attack faster and more effectively. Some bad actors will likely try to get around any cybersecurity boundaries the bots are taught, according to experts.
Adobe Launches Firefly Generative A.I., Which Lets Users Type to Edit Images
Adobe has launched an artificial intelligence tool called Firefly, which will allow users to make quick modifications to images by typing commands. Using generative AI, Firefly can generate different versions of an element in an image, change an image’s scene, and create a paintbrush based on something already in an image. Adobe said Firefly would prioritize giving creators opportunities to benefit from their skills and creativity and protect their work. The tool will initially launch as a private beta. The launch comes as AI investment accelerates, and ethical concerns have arisen around the propensity of models to generate inaccurate information and steal artist content.
For AI’s iPhone Moment, Nvidia Unveils a Large Language Model Chip
Nvidia has unveiled two new chips at its spring GTC conference, including the NVIDIA H100 NVL, which is designed to handle “inference” tasks on large language models, such as OpenAI’s ChatGPT. The chip is expected to be commercially available later this year and will offer faster inference performance at GPT-3 than the prior generation. The H100 NVL is expected to compete with Intel’s Xeon family, which has long dominated the inference tasks. Nvidia’s L4 for AI Video chip, which offers 120 times more AI-powered video performance than CPUs, will be integrated into systems from over 30 computer makers, including Google.
Society’s Technical Debt and Software’s Gutenberg Moment
Software production has long been hindered by its high cost and complexity, leading to a society-wide technical debt. However, recent developments in artificial intelligence, such as Large Language Models (LLMs) like ChatGPT, have made it easier to interact with conversational AI. While there is immense hype around the potential of software, the Baumol’s cost disease phenomenon shows that the cost of sectors with low productivity, such as software development, can grow exponentially compared to highly productive sectors. AI may be the solution to lower the cost of software production, but it could also lead to significant job losses and impact sectors, tax revenues, and even societal stability. In other words, the software industry may be eating the world, but it’s also eating itself.