Who is X’s iruletheworldmo?

Also: Perplexity’s popularity surges as AI search start-up takes on Google

Author

AI KATANA
August 11, 2024

Morning!

This week’s newsletter highlights key developments across the AI and tech sectors, from speculation around the mysterious X account @iruletheworldmo, which hints at upcoming OpenAI projects, to OpenAI’s concerns about its AI voice technology potentially fostering emotional bonds with users. We also explore Arati Prabhakar's influence on U.S. AI policy, Alibaba's breakthrough in AI math models with Qwen2-Math, and the rapid rise of Perplexity AI as it challenges Google in the search market. Additionally, the comparison between AI image generators Flux and Midjourney reveals Midjourney’s edge in human realism, while significant venture capital movements, including Thrive Capital's $5 billion raise and major funding in the travel tech sector, underscore ongoing investor confidence in AI-driven innovation.

Sliced:

  • ❓Who is X’s iruletheworldmo?

  • 👄 OpenAI worries its AI voice may charm users

  • 🇺🇸 Trailblazing Tech Leader Helps Shape U.S. AI Strategy

  • 📈 Perplexity’s popularity surges as AI search start-up takes on Google

  • 👩🏼‍💼 Four generative AI use cases for businesses

  • ➗ Alibaba claims no. 1 spot in AI math models with Qwen2-Math

In a recent development capturing the AI community's attention, the X account @iruletheworldmo has become a focal point of speculation and debate. With over 20K followers, this enigmatic account has been sharing cryptic tweets and memes about purported upcoming OpenAI projects, including a reasoning enhancement technique dubbed "Strawberry" (or Q*), a significantly upgraded GPT-4, and a potentially groundbreaking but controversial GPT-5 (internally called "GPT X"). The account also hints at safety concerns surrounding OpenAI's Sora and Voice technologies. While some believe @iruletheworldmo could be an OpenAI insider or even an advanced AI, others remain skeptical, viewing the claims as unsubstantiated hype. Adding intrigue to the situation, OpenAI CEO Sam Altman reportedly engaged with one of the account's tweets, further fueling speculation about its credibility. As with all unverified information, the AI community approaches these claims cautiously, recognizing the need for official confirmation before drawing conclusions.

OpenAI has raised concerns about the potential for its realistic AI voice feature to create emotional bonds between users and the AI, potentially at the expense of human interactions. The company highlighted risks associated with anthropomorphizing AI, where users might develop misplaced trust or dependency, which could impact social norms and human relationships. The voice capabilities of OpenAI's GPT-4 model, which enable more human-like interactions, have led to instances where testers expressed sentiments suggesting emotional connections with the AI. OpenAI plans to further investigate how extended use of these voice features might affect users' social behaviors and emotional attachment to the technology. Additionally, the ability of the AI to convincingly repeat false information has also been flagged as a concern, particularly given its recent controversy involving a voice similar to actress Scarlett Johansson's. OpenAI emphasizes that AI should not replace genuine human interaction and is continuing to study the broader implications of its voice technology on society.

Arati Prabhakar, the first woman and person of color to direct the U.S. White House Office of Science and Technology Policy (OSTP), has played a pivotal role in shaping the nation's AI strategy. Since her appointment in 2022, Prabhakar has advised President Biden on AI regulation, culminating in the Executive Order on the Safe, Secure, and Trustworthy Development and Use of AI. This order, signed in October 2022, sets forth critical objectives such as safeguarding consumer privacy, developing watermarking systems for AI-generated content, and addressing intellectual property concerns. Prabhakar's career, which spans leadership roles at NIST, DARPA, and the private sector, reflects her dedication to leveraging technology for societal good. Under her guidance, the U.S. has also led efforts at the United Nations to make AI regulation a global priority. Prabhakar's work continues to influence the responsible development of AI, positioning the U.S. as a leader in the ethical governance of this transformative technology.

Perplexity AI, a rapidly growing AI-powered search start-up, has seen its popularity surge, significantly increasing its monthly revenues and usage in 2024. Founded just months before the launch of ChatGPT by former Google intern Aravind Srinivas, Perplexity answered 250 million questions in July alone, nearly half its total for all of 2023. The company recently secured $250 million in funding, tripling its valuation to $3 billion, and is transitioning from a subscription-based model to an advertising-driven approach, bringing it into closer competition with Google. Despite its success, Perplexity has faced criticism for its data-gathering techniques, including accusations of plagiarism, which it has responded to by enhancing citation practices and launching a revenue-sharing program with publishers. Unlike its larger competitors, Perplexity licenses AI systems rather than developing its own models, and it aims to provide high-quality answers by focusing on reliable sources, such as journalism and academia. However, the introduction of ads has raised concerns about maintaining user trust. As AI-driven search continues to evolve, Perplexity remains a key player in the increasingly competitive market.

Generative AI is transforming business operations by enhancing productivity, reducing costs, and enabling new ways of working across various industries. Companies are leveraging AI-powered virtual assistants to improve customer interactions and automate routine tasks, freeing employees to focus on more strategic activities. Intelligent search tools, tailored with company-specific data, enable precise and contextually relevant information retrieval, boosting decision-making capabilities. Content summarization powered by AI significantly reduces the time required to process large volumes of information, benefiting sectors like healthcare and finance. Additionally, generative AI is revolutionizing document processing by automating and personalizing document creation, analysis, and management. These advancements are helping businesses gain a competitive edge, with those adopting generative AI early poised to lead in innovation and industry standards.

Alibaba has claimed the top spot in AI math models with the release of Qwen2-Math, a series of math-specific large language models (LLMs) developed by Alibaba Cloud. The most powerful variant, Qwen2-Math-72B Instruct, has achieved remarkable results on several math benchmarks, including an 84% score on the MATH Benchmark, 96.7% on the grade school math benchmark GSM8K, and 47.8% on the collegiate-level math benchmark. This places Qwen2-Math ahead of other leading AI models like OpenAI's GPT-4o and Google's Math-Gemini. Designed to address the traditionally weak performance of LLMs in solving mathematical problems, Qwen2-Math is positioned as a valuable tool for solving complex equations and supporting STEM fields. Although not fully open-source, Qwen2-Math is available under a permissive license for most commercial uses, making it accessible to startups and enterprises.

🧑🏽‍💻 AI Jobs

🛠️ AI tools updates

In a head-to-head comparison between Flux and Midjourney, two leading AI image generators, Midjourney emerged as the overall winner due to its superior ability to render realistic human textures, particularly in challenging scenarios such as capturing detailed skin textures and dynamic environments. The comparison involved generating images based on five different prompts, where Midjourney consistently produced more realistic characters and better structured scenes in most categories, despite Flux occasionally excelling in background details and overall image composition. The results highlight Midjourney's recent improvements in skin texture rendering with its latest version, v6.1, which gave it an edge in producing more lifelike images, while Flux, though strong in other aspects, fell slightly short in human realism.

💵 Venture Capital updates

Thrive Capital has successfully raised $5 billion for its largest-ever venture capital funds, reflecting investor optimism fueled by the ongoing AI boom. Founded by Josh Kushner, Thrive has gained prominence in the startup ecosystem, particularly through its close relationship with OpenAI, the company behind ChatGPT. Thrive's latest fundraising is divided into a $1 billion fund for early-stage ventures and a $4 billion fund for later-stage investments. The firm’s strategy, which focuses on AI and other technological innovations, has positioned it as a significant player in the venture capital industry, even as the broader startup sector faces challenges such as a sluggish IPO market and increased antitrust scrutiny. Thrive’s recent successes include major investments in OpenAI, Scale AI, and Stripe, as well as a diverse portfolio that spans technology, entertainment, and finance.

This week, travel startups raised nearly $240 million, with the majority of the funding going to Flyr, a modern retail platform for airlines, which secured $225 million in a Series D round. This funding round, led by WestCap and supported by BlackRock, the Abu Dhabi Investment Authority, and Avianca, marks the largest travel tech deal of the year. Flyr's platform uses AI to offer dynamic pricing and personalized offers to airlines, with plans to expand similar services to hotels. Other notable startups that raised funds include Sora Fuel, a developer of sustainable aviation fuel, which secured $6 million, and Wanderboat, an AI trip planner, which also raised $6 million. Berlin-based YGO Trips, a booking platform for vacation packages, completed the week's funding activity by raising $2.7 million. These investments highlight the ongoing interest and innovation in the travel tech sector, particularly in AI-driven and sustainable solutions.

🫡 Meme of the day

⭐️ Generative AI image of the day