Apple Intelligence

The AI World Today

• Apple Unveils iPhone 16 AI

• Musk Updates on Dojo Supercomputer

• Amazon Launches AI Voice Clones

• AI Ideas Rated More Novel

• Summit On AI Nuclear Control

  +

• Heads Up

• AI Solutions

Apple Introduces iPhone 16 and AI Integration

At its major product event, Apple introduced the iPhone 16, emphasizing its new AI-powered "Apple Intelligence" features. Designed “from the ground up” for AI, the iPhone 16 integrates tools like text generation and real-world object recognition via the device’s camera. Apple Intelligence allows users to edit text, create custom emojis, and search photos or messages with descriptions or voice commands. The iPhone 16 also features a more resilient display, an "action button" for customizable tasks, and the new A18 chip for improved performance. Alongside the iPhone, Apple unveiled updates to its AirPods and Apple Watch. These AI integrations mark Apple's deeper dive into the competitive AI race, following announcements made earlier this year at its Worldwide Developers Conference.

Musk Announces Dojo2 and 3 Coming in Next Two Years

Tesla CEO Elon Musk announced that the company’s Dojo 2 supercomputer is set for volume production by late 2025, with Dojo 3 expected by 2026. Speaking on the All-In podcast, Musk highlighted that while he has confidence in Dojo, its full potential will only become clear with the third iteration. Dojo plays a vital role in Tesla’s Full Self-Driving (FSD) technology, training the neural networks that power autonomous vehicles. Dojo is designed to handle massive AI training workloads, comparable to Nvidia's B200 systems, known for their high-performance AI processing capabilities. Tesla plans to invest over $1 billion in Dojo by 2024, while also continuing to rely on Nvidia, with a $3-4 billion spend planned for hardware purchases.

Audible Introduces AI Voice Replicas for Audiobooks

Amazon is launching a beta program for Audible narrators to create AI-generated voice clones, aiming to speed up audiobook production. The US-only test invites a small group of narrators to submit voice recordings for training their AI replicas, with plans to extend the program to authors, agents, and publishers later in 2024. Narrators will retain control over which projects use their AI voice and can edit pronunciation and pacing. They will be compensated via a "Royalty Share" model. Titles narrated by AI replicas will be labeled accordingly. This move follows Amazon's introduction of fully synthetic voices for Kindle Direct Publishing, which has sparked concerns about job opportunities for human narrators as virtual voices gain popularity.

Study Finds AI Generates Novel but Less Feasible Ideas

A recent study comparing AI-generated research ideas to those from human experts in natural language processing (NLP) found that AI ideas were judged as more novel, though slightly less feasible. Over 100 NLP researchers participated, writing and reviewing ideas without knowing whether they were AI- or human-generated. The study revealed AI-generated ideas were rated significantly higher for novelty but lower in feasibility. Challenges in AI idea generation included self-evaluation difficulties and a lack of diversity in outputs. This study underscores the potential of AI to spark innovation and contribute to scientific discovery, answering a key question about AI’s role in accelerating research. Despite some limitations, AI shows promise in driving novel ideas forward in scientific fields.

Global Summit Agrees on Human-Controlled Nuclear Weapons

At the Responsible AI in the Military Domain (REAIM) summit in Seoul, nearly 100 countries, including the U.S. and Ukraine, agreed that humans, not AI, should make key decisions regarding the use of nuclear weapons. The non-binding "Blueprint for Action" declaration emphasized the need for human control in nuclear weapons deployment and stated that AI applications in the military must adhere to national and international laws. The summit, co-hosted by Britain, the Netherlands, Singapore, and Kenya, did not secure China's signature, and Russia was excluded due to its invasion of Ukraine. The declaration recognized the rapid development of AI in military contexts, urging further discussions for establishing clear policies and procedures to regulate AI's military use.

Heads Up

AI Safety: US proposes requiring reporting for advanced AI, cloud providers

AI Funding: Australia's NEXTDC plans to raise $500 mln to tap AI boom via Asian expansion

AI Infrastructure : EU launches AI factories call to support developers

AI Startup: This German startup is Europe’s best hope for developing AI advancement outside Silicon Valley

AI Standards: Australia releases new mandatory guardrails and voluntary standards on AI – what you need to know

AI Solutions

iCardio.ai and SARC MedIQ Partner to Revolutionize Ultrasound Interpretation with AI

iCardio.ai, an AI startup backed by Cedars-Sinai, and SARC MedIQ, a leading provider of PACS (Picture Archiving and Communication System), have announced a strategic partnership aimed at advancing AI tools for ultrasound image interpretation. By integrating iCardio.ai's cutting-edge AI technology with SARC MedIQ's cloud-based PACS platform, the partnership will enhance efficiency, accuracy, and speed in diagnosing cardiovascular conditions. The collaboration seeks to revolutionize the healthcare industry by providing clinicians with automated AI insights that improve echocardiography quality and streamline ultrasound evaluation processes. Both companies aim to empower healthcare providers with advanced tools, ultimately improving patient outcomes. This partnership represents a significant leap forward in AI adoption within the cardiology sector, offering clinicians greater diagnostic capabilities through cloud-based solutions.