Perplexity has brought its Personal Computer assistant to Mac, extending its earlier Computer product from a cloud-managed agent into software that can work directly with local files, native apps and the web. The rollout begins for Perplexity Max subscribers, with the company also prioritizing people who previously joined the waitlist.

Inside the upgraded desktop app, Personal Computer lets users trigger tasks with text or voice and then execute them across Mac applications. Perplexity says it can draft emails, finish items in Notes, add calendar events, organize desktop folders and browse locally with Comet, making the assistant feel closer to an operating layer than a chatbot.

The pitch is not just convenience but persistence. Perplexity recommends a Mac mini because the system can stay on continuously for longer-running jobs, while users can initiate or manage work from an iPhone when away from their desks. That hybrid setup blends local access with server-side orchestration, aiming to improve privacy, responsiveness and continuity.

The broader significance is that AI assistants are moving beyond answering questions and into doing work across personal devices. By embedding orchestration into the desktop, Perplexity is pushing toward a future where consumers and professionals can hand off repetitive digital chores without fully surrendering control, especially when sensitive documents or app workflows need to remain on-device.

For users, the launch matters less as a flashy demo than as a sign of how premium AI products are being packaged: deeper system access, always-on availability and higher subscription tiers. Personal Computer is available now today on Mac for Max subscribers, broader Pro access is rolling out, and Windows support is planned next. That makes this release both a feature expansion and a strategic test of whether people will pay more upfront for assistants that can actively complete tasks instead of simply suggesting them.

Source: Perplexity

Robi’s Insights:

AI Beanie Can Read Your Thoughts: Silicon Valley startup unveils a smart beanie that converts your thoughts into text using AI-powered brain signals—no surgery required.

Source: Wired

The company continues to push forward on AI tools across text, images, and video, while rivals circle with their own increasingly polished demos. In other words, the race to generate ever more convincing synthetic reality is still very much on, just with slightly fewer familiar names attached to the pitch deck. Silicon Valley remains committed to its favorite tradition: building the future at breakneck speed and then swapping out the narrators mid-sentence, as if nobody will notice the plot is still being improvised in real time. And while leadership churn is practically a KPI at this point, it does little to answer the only question that matters: when these dazzling clips graduate from viral curiosities into tools people actually rely on without a disclaimer longer than the output on a daily basis.

Source: The Verge

Musk Turns xAI Into AI “Cloud Giant”: Elon Musk’s xAI plans to rent massive GPU power to Cursor for training AI models, signaling its shift into a cloud-style computing provider.

Source: Business Insider

A recent study reports that popular artificial intelligence chatbots provide inaccurate or incomplete answers to medical questions nearly half the time, raising concerns about their reliability for health information. Researchers evaluated responses from multiple widely used systems by asking standardized health questions and comparing answers against established clinical guidelines. The analysis found frequent omissions of key details, misleading statements, and inconsistent recommendations depending on how questions were phrased. While some responses were accurate, overall performance varied significantly, with certain tools performing better in general knowledge than in nuanced clinical decision support. The study emphasized that chatbots often lack context awareness and may generate confident sounding but incorrect advice. Experts caution that patients should not rely on these tools for diagnosis or treatment decisions and recommend consulting qualified healthcare professionals.

The findings highlight the broader challenge of integrating artificial intelligence into healthcare responsibly, especially as more people turn to digital platforms for quick answers. Ensuring that systems are trained on high quality medical data and regularly updated will be essential to minimize harm and build public trust in emerging technologies. Future research will likely explore ways to benchmark chatbot performance, incorporate clinician oversight, and develop regulatory standards to guide safe deployment across different healthcare settings worldwide global use.

Source: Dig Watch

AI Tool You Did Not Know You Needed

Can you outsmart AI?

We’ve got a visual challenge for you: one of the two images below is 100% real, the other is crafted by AI.

Click option below A or B. 👇

Ready to take your AI journey further?

Craving more AI chaos?

Read Daily AI News at BitBiased.AI. Support us by following us on LinkedIn and X ( Twitter ).

Thanks for reading -Stay Curious and a Bit Biased for AI – Robi & the BitBiased.AI team