AI This Week: Apple’s AI Rollout, OpenAI’s Voice Mode, and EU’s AI Act 

7 mins

This week’s AI news shows a blend of technological advancement and practical applications unfolding. Learn about the latest in AI, from Apple’s updates and OpenAI’s new features to a startup’s approach to AI companionship and the EU’s engagement with AI stakeholders. Plus, explore how AI is reshaping traditional revenue models in publishing.

Apple Introduces Limited AI Capabilities with New iOS Update

Apple has started integrating its much-anticipated AI suite, Apple Intelligence, into the iOS 18.1 developer beta, albeit with significant limitations. Initially unveiled at the Worldwide Developer Conference (WWDC) in June, the full spectrum of AI features was not included in the first rounds of both developer and public betas.

Currently, the AI functionalities are restricted to users with newer hardware, specifically the iPhone 15 Pro, iPhone 15 Pro Max, and devices equipped with M1 chips, provided the device language is set to U.S. English. The device region must also be set to the U.S., though this does not affect App Store settings. 

Interested users can join a waitlist via a new menu in the settings app labelled “Apple Intelligence & Siri.” Notifications will alert users when they can start using the features, which include downloading necessary AI models. However, full-feature access may take time.

Featured Image: Apple

Among the new features, Apple has enhanced Siri with fresh animations and the ability to respond to text prompts via double taps on the bottom bar. The AI suite also introduces advanced writing tools such as proofreading, tone adjustment (Friendly, Professional, and Concise), text transformation into lists or tables, and critical point highlighting.

The Mail app also benefits from a new summarization tool that provides one-line previews of emails and prioritizes important messages. Furthermore, a new “Reduce Interruptions” focus mode utilizes AI to curate essential notifications only.

Despite these introductions, many anticipated features like ChatGPT integration and the Image Playground are still under development, with bug fixes and optimizations ongoing, according to reports from Bloomberg. Apple’s cautious rollout indicates a deliberate approach to integrating
AI within its ecosystem to ensure stability and user satisfaction before fully deploying the entire suite of Apple Intelligence capabilities.

OpenAI’s New Developments in Voice Technology and AI-Enhanced Search

OpenAI is introducing significant updates to enhance user interactions and search capabilities. The first is the rollout of SearchGPT, a prototype that transforms how users retrieve information online. SearchGPT integrates AI’s conversational abilities with real-time web information to deliver fast and accurate answers, streamlining the web search experience. This prototype is part of a broader initiative to gather feedback from a limited user base and publishers, with plans to integrate the best features directly into ChatGPT in the future. 

Key features of SearchGPT include:

Direct Responses: It quickly answers queries with the latest web data, citing sources.
– Conversational Context: Users can ask follow-up questions in a natural, conversational style, maintaining context throughout the dialogue.
– Publisher Partnerships: OpenAI is working closely with publishers and creators to enhance how content is discovered and engaged with online. This includes giving publishers more control over how their content appears in SearchGPT.

Concurrently, OpenAI is gradually introducing an advanced voice mode for ChatGPT, currently available to a select group of ChatGPT Plus subscribers. Demonstrated initially at the GPT-4o launch event, this feature faced early feedback and safety concerns due to its likeness to Scarlett Johansson’s voice from the film Her. As a result, its release was delayed to refine safety features and ensure the voice mode met OpenAI’s quality standards. This enhanced voice mode allows for dynamic interactions like story narration and on-the-fly response adjustments and now employs four preset voices recorded with professional voice actors to avoid impersonating specific individuals or public figures. Additional safety measures have been implemented, including new filters to block the generation of copyrighted audio content.

Features Image: OpenAI

New AI Startup Shifts Focus from Productivity to Companionship

Avi Schiffmann’s latest project, “Friend,” is redefining digital companionship by offering an AI companion that focuses on emotional support rather than productivity. Unlike traditional AI tools, Friend is designed as a wearable pendant that actively participates in users’ daily lives, providing support and interaction whenever needed.

This innovation departs from conventional AI applications by prioritizing personal connection and companionship. Friend communicates through a mobile app, offering interactions that are meant to be temporary and personal, mirroring real-life conversations. Scheduled to launch in January, the first 30,000 units will be available at $99 each, without a subscription fee.

Featured Image: Friend

Schiffmann envisions Friend as an integral part of the user’s social circle, potentially becoming one of the “five people” one interacts with the most. However, he emphasizes that Friend is meant to enhance, not replace, human interaction.

With Friend, Schiffmann introduces a new dimension to AI, exploring how technology can foster meaningful digital relationships that complement our social lives.

EU Invites Stakeholder Input on AI Regulations

The European Union has begun a public consultation to establish rules for general-purpose AI models under its new AI Act. The Act, which will take effect on August 1, 2024, targets AI applications with a risk-based regulatory approach. These guidelines will affect major AI developers like Anthropic, Google, Microsoft, and OpenAI.

The consultation aims to engage a wide range of stakeholders, including AI providers in the EU, businesses, civil society, rights holders, and academic experts. It will focus on transparency, risk assessment, and intellectual property for AI systems, particularly those posing systemic risks.

An essential element of this initiative is crafting a summary template for AI model training content, which will be refined through virtual meetings and workshops involving various stakeholders. The AI Office encourages diverse participation to ensure the process is inclusive and transparent.

Stakeholders are invited to submit their input by September 10, 2024, as the EU strives to foster innovation while ensuring AI development adheres to rigorous regulatory standards.

Perplexity AI to Share Ad Revenue with Publishers

Perplexity AI has announced plans to share advertising revenue with news publishers when their content is referenced by its AI chatbot, addressing ongoing concerns about plagiarism and unethical content use. Dmitry Shevelenko, Perplexity’s head of business, disclosed to TechCrunch that discussions about this revenue-sharing model started even before accusations from publishers began. This strategy is part of a broader effort to sustain the news ecosystem essential for Perplexity’s continued provision of accurate information.

The company’s new initiative, the Perplexity Publishers Program, will share ad revenue and provide publishers with access to its APIs, developer support, and enhanced privacy features through the Perplexity Enterprise Pro offering. Initial partners include prominent publishers like Automattic, Der Spiegel, and TIME.

Featured Image: Perplexity

This move comes as generative AI alters how users search for information, prompting publishers to explore new monetization methods. Perplexity’s program is designed to align its interests with those of publishers, ensuring a diverse and vibrant media landscape without competing directly for revenues.

While the specifics of the revenue share have not been disclosed, Shevelenko mentioned it would be a “double-digit” percentage. The formal introduction of ads on Perplexity’s platform, featuring partnerships with top-tier brands across various industries, is slated for the coming months.

This approach marks a distinct shift from traditional tech platforms, as Perplexity aims to prioritize revenue sharing over traffic generation as the primary value offered to publishers. This initiative sets a precedent that could influence future collaborations between AI platforms and content creators.

Keep ahead of the curve – join our community today!

Follow us for the latest discoveries, innovations, and discussions that shape the world of artificial intelligence.