Anthropic has launched a significant upgrade to its Claude AI chatbot, allowing users to select from three distinct writing styles: concise, explanatory, and formal. This feature aims to tailor the chatbot’s responses to match individual user preferences, providing clearer and more polished communication or more direct answers as desired. Users can also create custom writing styles by uploading examples of their own writing. Currently, this feature is only available on the web version of Claude and is not yet accessible on mobile applications.
H, a Paris-based AI startup founded by former Google employees, has launched its first product, Runner H, after raising two hundred twenty million dollars in a seed round earlier this year. Despite losing three of its five co-founders due to operational disagreements, the company has developed Runner H as an “agentic” AI designed for businesses, focusing on quality assurance and process automation. With a proprietary model based on just two billion parameters, H aims to provide efficient solutions in areas like robotic process automation and business process outsourcing. The company is also raising a Series A funding round, emphasizing the capital-intensive nature of AI development. H claims its model outperforms competitors, including Anthropic and Meta, in efficiency and operational costs.
A new study from the Tow Center has raised alarms about OpenAI’s ChatGPT misrepresenting and misattributing content from publishers, including those in licensing agreements. Researchers tested ChatGPT’s search tool with twenty publishers and found that the chatbot frequently provided inaccurate citations, misquoting content, or failing to acknowledge its inability to provide correct citations. This inconsistency could harm publishers’ visibility and potentially encourage plagiarism. OpenAI responded by labeling the study as atypical, highlighting their commitment to improving citation accuracy while serving a user base of two hundred fifty million weekly.
Why do we care?
For businesses and IT providers, the key takeaway is the need to balance adoption of cutting-edge AI capabilities with a strong focus on reliability, transparency, and ethical considerations. Businesses should prioritize tools that align closely with operational needs while remaining user-friendly and secure. Smaller, targeted models like Runner H may offer a cost-effective path for specific use cases but should be scrutinized for scalability and robustness. Trust will remain a cornerstone of AI adoption. Providers should actively monitor and address the limitations of AI tools to ensure long-term value for clients.