All articles tagged with #genai
MIT SMR columnists Thomas Davenport and Randy Bean identify five key AI trends for 2026, including the deflation of the AI bubble impacting the economy, the rise of 'AI factories' for faster AI development, the shift of GenAI to an enterprise resource, the potential of agentic AI despite current overhype, and ongoing debates about AI management roles within organizations.
Rob Pike criticizes the use of AI-generated thank-you emails, viewing them as disrespectful and a marketing gimmick that undermines genuine human interaction. The article discusses broader concerns about AI's role in society, including environmental impact, intellectual property, and the authenticity of AI communication, emphasizing that AI tools lack agency and should not replace sincere human effort.
AI is being used to rapidly design thousands of potential antibiotics to combat rising antimicrobial resistance, with some candidates showing promise in early tests, but challenges remain in manufacturing and stability before they can be tested in humans.
A recent MIT study reveals that 95% of organizations investing heavily in AI see no return, raising concerns about the effectiveness of AI spending and its impact on investor confidence, especially amid volatile markets and high tech expenditures.
YouTube is introducing new GenAI-powered features to its Shorts platform, including photo-to-video creation, AI effects like underwater visuals, doodle transformations, and an AI Playground for generating videos and music from prompts, initially powered by Veo 2 and soon upgraded to Veo 3, available in select countries.
Scale AI is laying off 200 employees, about 14% of its workforce, as part of a restructuring to better align with market demands and reduce inefficiencies after rapid expansion in generative AI capacity. The company plans to focus on core areas like code, languages, and audio, while deprioritizing less promising projects, and aims to hire hundreds more in the second half of 2025.
Nvidia has partnered with Israeli startup RGo Robotics to enhance mobile robots with GenAI-powered autonomy. By integrating RGo’s perception engine with Nvidia’s compute and algorithms, the collaboration aims to improve robots' ability to perceive, understand, and interact with their environment, enabling them to perform complex tasks safely and productively. This partnership was announced by Nvidia CEO Jensen Huang at COMPUTEX in Taiwan, showcasing the deployment of the technology in real-world settings.
Keywords, a global service provider, conducted an experiment called 'Project Ava' to make a 2D video game using generative AI tools and technology, but found that while some AI tools can simplify or accelerate certain processes, they are currently "unable to replace talent." Despite this, the company is pushing ahead with other GenAI R&D projects and is leveraging AI tools in other ways, such as aiding customer support. Additionally, Keywords has recruited a former head of gaming from AWS Games to lead its AI Centre of Excellence and is continually "mapping the landscape" for AI tools that can be deployed in the game development cycle.
Anthropic, a startup backed by Google, has announced its latest AI chatbot models, Claude 3, claiming to outperform OpenAI's GPT-4. Claude 3 is the company's first multimodal GenAI, capable of analyzing both text and images, with enhanced performance on specific benchmarks. It can process multiple images in a single request and is designed to better follow multi-step instructions and produce structured output. However, it has limitations in image processing and is prone to mistakes with low-quality images. Anthropic plans to release enhancements to Claude 3 in the coming months and aims to raise significant capital to remain competitive in the AI space.
Large Language Models (LLMs) are rapidly evolving, with Meta's Code Llama 70B and other specialized LLMs tailored for code generation gaining attention. While concerns about LLMs replacing human programmers persist, the focus should be on selecting the right LLM for specific use cases. Smaller, more focused LLMs are gaining momentum, offering safer and more relevant options for AI-assisted code generation. However, the likelihood of GenAI tools replacing human programmers entirely remains uncertain, as the human element in software development, including design, innovation, and quality control, remains crucial.
Apple CEO Tim Cook announced during the annual shareholders meeting that the company will "break new ground" in GenAI this year, following reports of reassigning staff from a scrapped EV project to work on GenAI initiatives. Apple has been slow to invest in GenAI compared to its rivals, but is reportedly planning to upgrade Siri, Spotlight, Keynote, and Apple Music with GenAI models, with potential new features expected to be showcased at the Worldwide Developer Conference. The company has also been intensifying its focus on GenAI, publishing open source models and tools, investing $1 billion a year in GenAI efforts, and rumored to be upgrading the Neural Engine in upcoming iPhone 16 models.
Adobe has unveiled Project Music GenAI Control, a platform that uses AI to generate audio from text descriptions or a reference melody and allows users to customize the results. The tool, developed in collaboration with researchers at the University of California and Carnegie Mellon, is still in the research stage and does not have a UI yet. However, it raises ethical and legal concerns regarding AI-created music and copyright issues. Adobe is working on watermarking technology to help identify audio produced by the tool and takes a responsible approach to developing GenAI tools against data that's under license or in the public domain to avoid potential IP issues.
Orbital Materials, a startup founded by former DeepMind researcher Jonathan Godwin, is developing GenAI technology to support the discovery and manufacturing of new physical materials. Their AI-powered platform, Linus, uses natural language instructions to generate 3D molecular structures that meet specific criteria, aiming to accelerate the materials design process. With $16 million raised in a Series A round, the company plans to expand its data science and wet lab teams to bring advanced materials to the proof of concept or pilot demonstration phase and seek outside manufacturers as partners.
Google has rebranded its Bard chatbot to Gemini, offering an Android app that can replace Google Assistant with the GenAI-powered assistant, providing expanded capabilities such as generating workout plans and writing emails. The Gemini app is available for all Android phones, and the process to replace Google Assistant involves downloading the Gemini app, agreeing to terms and conditions, and activating Gemini. If unsatisfied, users can easily switch back to Google Assistant via the Google app.
Google's Gemini chatbot, a competitor to OpenAI's ChatGPT and Microsoft's Copilot, has received mixed reviews in its performance tests. While it excels in research and productivity, it falters in some areas. The chatbot offers a lightweight version, Gemini Pro, for non-paying users and a more powerful model, Gemini Ultra, behind a paywall. Testing Gemini Ultra with a set of questions revealed its thorough but sometimes overly detailed responses, reluctance to provide certain answers, and limitations in certain functionalities. Despite potential for improvement, the $20 per month price tag may not currently justify its capabilities compared to other similar models.