Hello again AI fans!

This week’s newsletter covers Apple's strategic partnership with Alibaba to integrate advanced AI features into iPhones sold in China, addressing local regulatory challenges and revitalising its market position amidst declining sales. Meanwhile, Moonshot AI's Kimi k1.5 is disrupting the global AI landscape with superior multimodal capabilities, outperforming leading models like GPT-4o in benchmarks for mathematics, coding, and reasoning.

This week’s edition:

• Apple Expands AI Strategy in China with Dual Partnerships

• Kimi K1.5: China's New AI Challenger

• Your 5-Min AI Learning: Take Your AI Assistants Offline with Jan

• New AI Tools

• More AI News

• AI Image of the Week

Read time: 8 minutes

LATEST NEWS

APPLE PARTNERS WITH CHINA

Apple Expands AI Strategy in China with Dual Partnerships

What You Need to Know

Apple is strengthening its AI capabilities in China through partnerships with state-backed firms Primavera Capital Group and Hillhouse Capital. This move aims to enhance Apple's AI offerings while navigating the complex regulatory landscape in China. The collaboration focuses on developing AI models that comply with Chinese regulations and cater to the local market's needs.

The Finer Details

• Apple is partnering with Primavera Capital Group and Hillhouse Capital

• The focus is on developing AI models tailored for the Chinese market

• The strategy aims to comply with China's strict data and AI regulations

• Apple plans to invest in Chinese AI startups and research institutions

• The partnerships will help Apple compete with local tech giants like Baidu and Alibaba

So What?

This strategic move allows Apple to expand its AI presence in China while adhering to local regulations. By collaborating with state-backed firms, Apple can potentially gain favour with Chinese authorities and access the vast Chinese market more effectively. This approach could give Apple a competitive edge in AI development and implementation within China, potentially leading to innovative products and services tailored for Chinese consumers.

MOONSHOT AI

Kimi K1.5: China's New AI Challenger

What You Need to Know

Kimi K1.5, developed by Chinese tech company Moonshot AI, is a large language model (LLM) that's making waves in the AI world, signalling a potentially transformative moment in artificial intelligence development. As a sophisticated AI system, it's being touted as a potential rival to OpenAI's GPT-4, showcasing China's rapidly advancing capabilities in cutting-edge technological innovation. The model has demonstrated impressive performance across various benchmarks, particularly in Chinese language tasks, and has caught the attention of both domestic and international tech communities, sparking discussions about the evolving landscape of global AI research.

The Finer Details

• Developed by Moonshot AI, a Beijing-based startup

• Outperforms GPT-4 in some Chinese language tasks

• Shows strong capabilities in English language processing

• Trained on a diverse dataset, including scientific papers and code

• Raises questions about data sources and potential biases

• Faces challenges in global adoption due to geopolitical tensions

So What? 

Kimi K1.5 represents a significant step forward for China's AI industry, potentially narrowing the gap with Western AI giants and challenging the current technological hegemony. Its emergence underscores the intensifying global AI race, highlighting the critical importance of continuous innovation and strategic technological development. The model's potential global impact may be constrained by ongoing concerns over data privacy, potential censorship, and complex international tensions surrounding Chinese technological exports, yet it nonetheless represents an important milestone in AI's global evolution.

5 MINUTE AI LEARNING

YOUR 5 MIN AI LEARNING

Take Your AI Assistants Offline with Jan

This week in our AI learning journey, we're diving into Jan, an incredible open-source ChatGPT alternative that's changing the game for privacy-conscious tech lovers. Jan runs completely offline on your computer, offering customisable AI assistants that can boost your productivity tenfold. With global hotkeys, in-line AI capabilities, and an OpenAI-equivalent API server right on your localhost, this tool is perfect for those who want more control and security over their AI experience. 

Additionally,  with a mobile app on the horizon promising seamless workflow integration, Jan is definitely a tool you'll want to explore. Try it out when you next need a powerful, privacy-first AI companion that puts you in the driver's seat!

LATEST AI TOOLS

Riffusion Fuzz 

Generate realistic sounding music from a prompt (Free)

“The more you use Riffusion, the more the model adapts to your unique taste and aesthetic. You can create and listen to music made with our latest foundation model, FUZZ. We see FUZZ as a new instrument that can create full-length, high quality songs from text, audio, and visual prompts.”

Cliprun 

A tool that can run Python code in a browser (Freemium)

“Execute code from ChatGPT, Claude, or GitHub with a simple right-click. No environment setup required. Analyze data directly in your browser. Visualize data with matplotlib, seaborn, and plotly output.”

WhisperTranscribe 

A tool that converts audio into text- available in various languages (Paid)

“Generate a transcript, a summary, show notes, titles, social media posts, a blog post and more. Your personal ChatGPT for your audio and video. 95% accurate transcripts in 55+ languages”

AI HIGHLIGHTS

Tensions over AI governance highlighted at Paris Summit

The AI Action Summit in Paris this week exposed tensions between China and Western nations over AI safety and collaboration, with China's Fu Ying criticizing US-China conflicts for hindering progress and advocating for open-source models. Yoshua Bengio highlighted risks from AI agents and debated the regulatory challenges of open-source versus closed-source systems, while Google DeepMind's Demis Hassabis emphasised balancing risk mitigation with market demands. The discussions underscored the challenges of aligning global cooperation, ethical concerns, and commercial interests in AI development.

News organisations, including Condé Nast, sue AI firm Cohere for copyright violations and producing "hallucinated" content

A group of major publishers, including Condé Nast and The Guardian, has sued AI firm Cohere for allegedly copying over 4,000 articles without permission to train its AI models and creating outputs that infringe on copyrights and trademarks. The lawsuit also accuses Cohere of bypassing paywalls, producing verbatim copies of content, and generating fake news under publishers' names, damaging their reputations. The plaintiffs seek damages and a permanent injunction to protect their intellectual property and ensure responsible AI innovation.

Google defends ending AI ethics pledges and scaling back DEI initiatives in staff meeting

Google executives defended the company's decision to end diversity, equity, and inclusion (DEI) initiatives and drop its commitment against weaponised AI during an all-staff meeting, citing compliance with federal directives and evolving legal requirements. The company emphasised the need to align its policies with changing global dialogues and maintained that its technology's benefits outweigh associated risks. Employees expressed concerns about these changes, raising numerous questions about the implications for Google's values and ethical stance.

DWP discloses use of AI tool to assist in benefit eligibility decisions

The DWP is using AI to help agents decide on ESA benefit applications by matching claimant health conditions to a central list, claiming an 87% accuracy rate and significant time savings. While in use since 2020, details were only recently shared, raising concerns about potential bias and the meaningfulness of human oversight. The DWP has resisted calls to publish a full list of AI tools used, despite transparency requirements.

AI IMAGE OF THE WEEK

For this week’s AI image, we at MegaSuperAI were inspired by the recent launch of Hello Kitty Island Adventure on PC. Initially a humorous jab in South Park, the concept has blossomed into a delightful game that invites players into a charming and whimsical world. With its cutesy adventures and engaging gameplay, we couldn’t help but wonder how an AI image generator would envision this vibrant universe.

We’re excited to see how an AI might capture the playful essence and enchanting landscapes of Hello Kitty Island Adventure!

The prompt used to generate this image on copilot: 

“Create a vibrant scene from Hello Kitty Island Adventure, featuring Hello Kitty and her friends exploring a colorful, whimsical island. Show them engaging in fun activities like building sandcastles, picking flowers, and enjoying a picnic under a bright blue sky. Include lush greenery, cheerful decorations, and playful animals to capture the essence of this enchanting adventure.” 

Send us your AI generated art for a chance to be featured.

FEEDBACK

• Let us know how we did - shoot us a quick email. We read every single one (no AI in sight!).

• If you enjoyed this newsletter please share with an awesome friend or two.

• If you are that awesome friend then hit the below button: