Today in AI | Grok needs a tutor | Google AI mode now smarter | Perplexity launches Comet Plus

The world of Artificial Intelligence has only begun to affect human lives. In times like these, staying up-to-date with the AI world is of utmost importance. Storyboard18 brings you the top AI news of the day.

Elon Musk’s xAI seeks ‘video game tutors’ to train Grok chatbot

Elon Musk’s artificial intelligence company xAI is hiring “video game tutors” to help its chatbot Grok learn the mechanics of game creation.

According to a job listing on xAI’s careers page, the company is seeking specialists who can train Grok in game mechanics, design, storytelling and user experience. The aim is to move Grok beyond text-based conversations and make it capable of generating playable, engaging video games.

Successful candidates will be responsible for providing detailed annotations, curated datasets and feedback on Grok’s attempts at producing game content. They will assess its work on creativity, technical performance and overall player experience, refining outputs to ensure both functionality and entertainment value.

The role requires prior experience in AI-assisted game development, playtesting or work with indie studios, alongside a strong personal interest in gaming. Tutors will collaborate with xAI’s technical teams, offering input on workflows and training tools.

Positions are available in Palo Alto, California, with a five-day in-office schedule, though remote work is permitted. Remote staff must complete two weeks on a fixed 9:00 am to 5:30 pm PST schedule before shifting to their local time zones. Applicants must use either a Chromebook, a Mac with macOS 11 or newer, or a Windows 10 computer, as well as a smartphone.

The role excludes applicants from Wyoming and Illinois, and no visa sponsorship is available. Pay ranges from $45 to $100 per hour, with benefits dependent on the applicant’s country of residence. Full-time positions may include medical cover, while part-time posts will not carry benefits.

Google AI Mode adds visual search for smarter, conversational shopping

Google has upgraded its AI Mode search to include conversational visual search, allowing users to shop using natural language and reference images rather than relying on traditional filters. The update, rolling out this week to US users in English, enables shoppers to describe products “like the way you'd talk to a friend,” according to Robby Stein, vice president of product for Google Search.

Users can now search with vague descriptions such as “barrel jeans that aren’t too baggy” and refine results with follow-ups like “show me acid-washed denim” or “I want more ankle length.” Searches can be initiated using uploaded reference images, photos taken on the spot, or a combination of images and text.

The new capability uses Google’s “visual search fan-out” technique, which runs multiple background queries to identify subtle details and secondary objects in images. Powered by Gemini 2.5’s multimodal capabilities and Google’s Shopping Graph of over 50 billion product listings, AI Mode delivers more relevant, shoppable results, including direct links to retailers, reviews, deals, and availability. Google refreshes more than two billion product listings hourly to maintain accuracy.

Beyond shopping, the visual search feature can be used for broader exploration, including interior design inspiration. Mobile users can query specific elements within images and receive conversational follow-ups about what they see. Stein admitted that AI Mode’s previous text-heavy responses to image searches could be “silly,” prompting this more visually intuitive approach, which builds on Google Lens and Image search technologies.

The rollout is expected to take several days to reach all users, marking the latest enhancement since AI Mode became broadly available in the US in March.

Delhi HC protects Sri Sri Ravi Shankar from AI deepfake misuse

The Delhi High Court has issued an interim order protecting spiritual leader Sri Sri Ravi Shankar from the unauthorized creation and circulation of deepfakes and AI-generated content using his name, voice, image or unique style of discourse, Bar and Bench reported.

Justice Manmeet Pritam Singh Arora restrained unknown parties, referred to as John Doe defendants, from exploiting Ravi Shankar’s persona for any commercial or personal purpose without consent, citing the risk of irreparable harm to his reputation and teachings.

The order comes after Ravi Shankar approached the Court following the emergence of fabricated videos between July and August 2025. These videos allegedly depicted him endorsing questionable health remedies and attributing false scientific claims or miraculous cures to him, misrepresenting his teachings and misleading the public.

The Court specifically barred defendants from using his name, voice, likeness, image, or any identifiable attributes, including his distinctive style of speech and delivery, in any AI-generated content. The injunction underscores the growing legal scrutiny over AI-enabled deepfakes, particularly when they threaten the reputation of public figures and spiritual leaders.

As AI technologies become increasingly accessible, the judgment highlights the urgent need for regulations and safeguards around the creation of digital replicas of real individuals, particularly public and spiritual figures whose teachings and credibility could be easily misrepresented.

The matter is expected to be further heard in due course to determine long-term remedies and damages for the misuse.

Perplexity AI launches Comet Plus; aims to deliver premium journalism in AI era

Perplexity AI has unveiled Comet Plus, a new subscription model designed to deliver high-quality journalism to users while compensating publishers for the ways their content drives value in the age of artificial intelligence.

Comet Plus, priced at $5 per month, is included at no additional cost for Perplexity Pro and Max subscribers. The model provides users and their AI assistants with direct access to reporting from trusted sources, aiming to combat clickbait and low-quality content that has dominated the internet for years.

The launch features a roster of partner publishers, including CNN, Fortune, Le Monde, Le Figaro, The Los Angeles Times, The Washington Post and Condé Nast publications such as The New Yorker, Wired, GQ, Vogue and Epicurious.

Under Comet Plus, participating publishers are compensated based on human and AI-driven engagement with their content, shifting away from the traditional “pageview-at-any-cost” model. The approach emphasizes trust, accuracy, and curiosity, while allowing publishers to maintain a direct relationship with their audience.

The move comes amid growing debates about how AI interacts with journalistic content and the challenge of ensuring that creators are fairly compensated as AI tools increasingly rely on online media for training and output.

With Comet Plus, Perplexity aims to create an ecosystem where premium journalism drives better AI responses, encouraging readers to ask more meaningful questions while ensuring that publishers and journalists are rewarded in the evolving digital landscape.