|
Anthropic explains how Claude's invisible text watermarks will work
Anthropic has clarified how it plans to apply invisible watermarks to Claude-generated text to comply with Europe's AI transparency rules. The system is based on Google DeepMind's open-source SynthID-Text approach, which creates detectable patterns using wording probabilities. The announcement builds on Anthropic's earlier commitment to text marking for regulatory compliance.
|
|
Anthropic's annualized revenue surges to $6.5B
Anthropic has seen its annualized revenue surge dramatically, adding $1.8 billion in just two months. The rapid growth underscores the company's rising dominance in the AI model market. The milestone reflects strong enterprise and developer adoption of its Claude models.
|
|
Groq raises $350M to fuel its pivot from AI chips to neocloud
Groq has raised $350 million at a $3.5 billion valuation as it shifts its business model from designing AI chips to operating a neocloud service powered by Nvidia hardware. The pivot marks a significant strategic change for a company once known for its custom LPU chips. The funding will support data center expansion and broader cloud service offerings.
|
|
Wispr raises $280M at $2B valuation as it looks beyond dictation
AI dictation startup Wispr has raised $280 million at a $2 billion valuation, signaling ambitions to expand well beyond voice-to-text. The company is moving into new productivity areas, including a newly launched AI meeting note-taker tool. The funding round positions Wispr as a broader AI productivity platform.
|
|
AI automation startup Relay shuts down, staff joins Google's Chrome team
AI automation startup Relay is shutting down, with its team set to join Google's Chrome division. Relay founder and CEO Jacob Bank hinted at ambitious plans to integrate AI into Chrome to help users get things done. The acqui-hire suggests Google is doubling down on AI-powered browser capabilities.
|
|
Nvidia investing $1.5B in SoftBank data center developer behind OpenAI project
Nvidia is investing $1.5 billion in SoftBank's data center development arm, ensuring its chips will power an upcoming OpenAI data center. The deal deepens ties between Nvidia, SoftBank, and OpenAI in the race to build AI infrastructure. The investment signals continued consolidation among major players in AI compute.
|
|
Amazon, which started off selling books, is destroying rare texts to train AI
Amazon is reportedly destroying rare and out-of-print books in order to use them as training data for its AI models. These texts are especially valuable for LLM training since they are not available online and represent knowledge not yet captured in existing datasets. The practice has raised ethical and cultural preservation concerns.
|
|
Whisker's AI-powered litter robot thinks my cats swapped bodies
Whisker's Litter-Robot 5 Pro uses AI to monitor cats' litter box habits and flag potential health issues, but the system struggles to reliably distinguish between multiple cats in the same household. A reviewer found that the robot's AI often confused their two cats, undermining the health-tracking feature. Despite the limitation, the device remains a convenient option for cat owners seeking automated litter management.
|