The SEC Newgate AI Weekly

As the AI landscape continues to blaze with innovation, another week packed with announcements is upon us.

Unveiling xAI: Elon Musk Ignites the Future with His Latest Venture

Another week another Musk headline.  Elon Musk has just unleashed xAI, an audacious AI venture that seeks to unlock the mysteries of the universe. The team comprises experts from DeepMind, OpenAI, Google, Twitter, and Tesla, who have worked on projects like AlphaCode and GPT chatbots. xAI is set to rival industry giants like OpenAI and Google in the chatbot space. With a focus on deep learning mathematics, xAI aims to take AI to the next level by developing a theory of everything for neural networks. However, Musk has already faced resistance from numerous academics and ethicists who argue that excessive attention on the expanding power and future risks of AI diverts us from the immediate harm inflicted by algorithms on marginalized communities.

Racing to the Top: Meta Introduces New AI Model to Outsmart Competitors

Meta is gearing up to unveil a new artificial intelligence model that can help start-ups and businesses create their own software solutions. They’re stepping up to compete head-on with industry giants like Microsoft-backed OpenAI and Google, who are currently leading the charge in the race to advance generative AI. The software, which can produce text, images and code, is driven by massive language models (LLMs) that learn from enormous amounts of data and need colossal computing power. Meta has already shared its own language model, dubbed LLaMA, with researchers and academics earlier this year, but the new version will be more accessible and adaptable by companies widespread. With the long-awaited release finally imminent, the future of AI is about to take another big leap forward.

Chatbot Clash

Anthropic, an AI startup, has launched Claude 2, an improved chatbot that aims to rival OpenAI's popular ChatGPT. The new chatbot is capable of analysing entire books and engaging in conversations for extended periods. It claims to be two times safer than its predecessor and incorporates trust and safety practices from reputable organizations. Anthropic trained Claude 2 using various documents, including the UN Declaration of Human Rights and Apple's terms of service. The chatbot has expanded capabilities, such as digesting technical documents or novels, writing longer texts, and fetching up-to-date information from the web. Additionally, it has shown improvements in coding and math problem-solving abilities and could be another new look into our AI-future.

Data Theft Battle

In other news this week, a lawsuit filed against Google accuses the tech giant of scraping user data without consent and violating copyright laws to train its AI products. The proposed class action suit targets Google, its parent company Alphabet, and Google's AI subsidiary DeepMind. The complaint alleges that Google has been secretly taking data from millions of Americans and using it to develop AI, including its chatbot Bard. Google's general counsel calls the claims baseless and states that the company uses publicly available data responsibly. The lawsuit seeks injunctive relief, damages, and compensation for individuals whose data was allegedly misappropriated by Google, with a request to allow users to opt out of data usage for AI training.

New Era of AI Governance: China's Landmark Regulations Impact ChatGPT and Beyond

Chinese regulators have finalized new rules that will govern generative AI, signalling a major leap in their oversight of this rapidly advancing technology. The new regulations, effective from August 15, specifically target AI services available to the public, requiring providers to secure licenses, conduct robust security assessments, and safeguard user information. Striving to strike a delicate balance between propelling AI leadership and cautious development, regulators aim to foster innovation and bolster the underlying infrastructure, to propel the future of generative AI in China.

And lastly, everything you need to know about Bard’s newest update:

• Bard’s reach expands with over 40 new languages and wider availability in all 27 EU countries and Brazil.
• Google Lens integration allows users to upload images and enhance their creativity during conversations with Bard, starting with English support
• Text-to-speech functionality is introduced in over 40 languages, including Hindi, Spanish, and US English, enabling users to listen to responses for a fresh perspective.
• Pinned and recent threads feature lets users organize and pick up where they left off in past Bard conversations, fostering an uninterrupted creative process.
• Sharing Bard conversations becomes simpler, allowing users to share specific parts or the entire chat via shareable links. This feature enables the inspiration of others, promotes creativity, and showcases collaborative processes. Additionally, users can modify Bard's responses to align with their desired creations. Lastly, exporting Python code to Replit is now possible, expanding Bard's code export capabilities beyond Google Colab and streamlining workflow for programming tasks.