ByteDance is developing a massive AI model with 10 trillion parameters to rival Anthropic's Mythos system, expected to be used for natural language processing and computer vision tasks. This move is part of ByteDance's efforts to develop AI technology that can compete with top US labs.
The model is nearing the scale of Anthropic's Mythos system and is expected to be three times larger than Moonshot's Kimi K3, a rival AI model. Meanwhile, CarMax has deployed conversational AI from Sierra for inbound sales calls, resolving customer queries and routing calls to human representatives when needed.
In the EU, the AI Act will require companies to be transparent about their AI systems, with experts predicting that corrective orders will be more common than large fines in the first year of enforcement. However, many companies struggle to see a return on investment in AI, despite significant investments.
China's AI ecosystem is not as open as it claims, posing challenges to the development of AI globally. In contrast, Michigan State University is using data and AI to develop mobility solutions, including autonomous vehicles and improving mobility for people with disabilities.
Companies need to manage their AI spending, prioritizing value over volume and tying spending to business outcomes. Cloudflare is addressing the shift to the Agentic Internet, where automated agents seamlessly integrate with human sessions. However, video language models struggle with event bookkeeping, even in simple tasks.
Key Takeaways
• ByteDance is developing a massive AI model with 10 trillion parameters to rival Anthropic's Mythos system.• CarMax has deployed conversational AI from Sierra for inbound sales calls.
• The EU AI Act will require companies to be transparent about their AI systems.
• Many companies struggle to see a return on investment in AI.
• China's AI ecosystem is not as open as it claims.
• MSU is using data and AI to develop mobility solutions.
• Companies need to manage their AI spending and prioritize value over volume.
• Cloudflare is addressing the shift to the Agentic Internet.
• Video language models struggle with event bookkeeping, even in simple tasks.
• Anthropic's Mythos system is a benchmark for AI models.
ByteDance trains massive AI model to rival Anthropic
ByteDance is developing a massive AI model with 10 trillion parameters to rival Anthropic's Mythos system. The model is expected to be used for natural language processing and computer vision tasks. This move is part of ByteDance's efforts to develop AI technology that can compete with top US labs. The model is still in the early stages of training, which typically takes several months to complete.
ByteDance targets mega AI model nearing Anthropic's Mythos
ByteDance is training a massive AI model that is nearing the scale of Anthropic's Mythos system. The model is expected to be three times larger than Moonshot's Kimi K3, a rival AI model. ByteDance's AI research team has been working on the model for several months and expects it to be completed soon.
What the first year of EU AI Act transparency enforcement could look like
The EU AI Act will require companies to be transparent about their AI systems. Experts predict that corrective orders will be more common than large fines in the first year of enforcement. Companies that make a genuine effort to comply will be treated more leniently. The Act will apply to AI systems that interact with humans, including chatbots and virtual assistants.
CarMax deploys conversational AI for customer calls
CarMax has deployed Sierra's conversational AI for inbound sales calls. The AI agent resolves customer queries and routes calls to human representatives when needed. CarMax reports higher call resolution and fewer unresolved calls since deploying the AI.
China's AI ecosystem is not as open as it claims
China's AI ecosystem is not as open as it claims to be. While China has made significant progress in AI development, its AI ecosystem is not as open as other countries. The article discusses the limitations of China's AI ecosystem and the challenges it poses to the development of AI globally.
Many companies struggle to see ROI on AI investments
Many companies are struggling to see a return on investment in AI. Despite significant investments in AI, many companies are not seeing the expected benefits. The article discusses the challenges of measuring ROI in AI and the need for companies to rethink their AI strategies.
Chatbot is just a guy answering questions
A chatbot called ChatTJB has been created that is simply a human, Tucker Bryant, answering questions. The chatbot is a commentary on how people interact with technology and how that affects relationships. Bryant hopes the chatbot will make people think about their use of technology.
MSU uses data and AI to shape mobility solutions
Michigan State University is using data and AI to develop mobility solutions. Researchers at MSU are working on interdisciplinary projects to advance mobility innovation. The projects include developing autonomous vehicles and improving mobility for people with disabilities.
Companies need to manage AI spending
Companies are struggling to manage their AI spending. AI budgets can quickly get out of control, and companies need to prioritize value over volume. The article discusses strategies for managing AI spending, including setting quarterly reviews and tying spending to business outcomes.
Cloudflare navigates the Agentic Internet
Cloudflare is addressing the shift to the Agentic Internet, where automated agents seamlessly integrate with human sessions. The company is focusing on assessing continuous actions to build trust and navigating the complexity of hybrid sessions.
Video language models struggle with event bookkeeping
Video language models struggle with event bookkeeping, even in simple tasks. Researchers have developed a new method to evaluate event counting in videos, and the results show that existing models have significant limitations.
Sources
- ByteDance trains massive AI model in bid to rival Anthropic
- ByteDance targets mega AI model nearing Anthropic’s Mythos
- What the first year of EU AI Act transparency enforcement could look like
- Sierra Lands CarMax as Conversational AI's Enterprise Land Grab Widens
- China’s AI ecosystem is not as open as it claims. Nor is any other country’s
- Many Companies Still Have Little to Show for Their AI Investments
- The Hottest New AI Chatbot Is Just a Guy Answering Your Questions
- How MSU Mobility experts use data, gen AI and policy analysis to shape advancements
- Headcount Isn’t Unlimited. AI Budgets Shouldn’t Be Either.
- Cloudflare: Navigating the Agentic Internet
- The Low Frequency Trap: Video Language Models Fail at Simple Event Bookkeeping
Comments
Please log in to post a comment.