• AI KATANA
  • Posts
  • Fujitsu taps generative AI to help speed drug development

Fujitsu taps generative AI to help speed drug development

Also: Financial Times tests an AI chatbot trained on decades of its own articles

Hi!

In today's newsletter, we delve into diverse developments across the AI spectrum: Fujitsu partners with Riken, employing generative AI to expedite drug development, potentially reshaping how diseases are treated with a significant reduction in time and resources. Meanwhile, Tokyo introduces an AI-based dating app to combat declining marriage and birth rates, embodying an innovative governmental approach to demographic challenges. In corporate shifts, Stability AI's founder Emad Mostaque steps down to explore decentralized AI, signaling a strategic pivot amidst fundraising hurdles. Google's introduction of the VLOGGER AI model stirs both excitement and ethical concerns over its potential to generate realistic video avatars. The Financial Times experiments with an AI chatbot, Ask FT, providing subscribers with information pulled from its extensive journalistic archives. A study reveals Generation Z's ambivalent feelings towards AI in the workplace, reflecting a blend of guilt, reliance, and optimism. WhatsApp announces AI-enhanced photo editing tools, promising to enrich user interactions without extra costs. Lastly, Saudi Arabia is reportedly in talks with Andreessen Horowitz to establish a $40 billion AI investment fund, indicating a significant stride towards AI and diversification from oil dependence.

Sliced:

  • 💊 Fujitsu taps generative AI to help speed drug development

  • 👰🏻‍♀️ Tokyo hopes AI app will spur marriages

  • 👀 Stability AI founder steps down to pursue 'decentralised AI’

  • 👱🏻‍♂️ Google's VLOGGER AI model can generate video avatars from images - what could go wrong?

  • 🤖 Financial Times tests an AI chatbot trained on decades of its own articles

  • 😕 36% Of Gen Z Feel Guilty About Using ChatGPT And AI At Work

Fujitsu, in collaboration with Riken, is harnessing generative AI to transform drug development by predicting the behaviors of target proteins with over tenfold speed compared to traditional methods. This innovative technology utilizes advanced cryo-electron microscopy images to create dynamic three-dimensional structures of proteins, aiding in understanding how substances interact with proteins to treat diseases. Initially set for trials in fiscal 2024, this approach significantly reduces the time required for drug development stages, such as determining the dynamic shapes and movements of relevant proteins, from a day to just two hours. The initiative is part of the Life Intelligence Consortium, engaging pharmaceutical companies and academic institutions to refine and leverage this technology. By providing detailed analysis at the early stages, it aims to concentrate resources on viable drug candidates, thereby increasing development efficiency and expediting the availability of treatments for infectious diseases and other conditions.

In an effort to counteract declining marriage and birth rates, Tokyo has launched a unique initiative by developing an AI-powered dating app aimed at spurring marriages. This move is part of the government's long-standing efforts, which include organizing matchmaking events and providing dating advice. The new app, set to release this spring, will delve into users' preferences and behaviors by asking over 100 questions to suggest potential matches based on data from 150,000 couples. To ensure safety and authenticity, the app will include an online interview process and require proof of single status and income. This initiative reflects the government's strategy to provide a starting point for individuals interested in marriage but uncertain about where to begin. The app targets residents, workers, or students in Tokyo, embodying the city's commitment to addressing demographic challenges by nearly doubling the budget for marriage-support services.

Emad Mostaque, the founder and CEO of Stability AI, has resigned to focus on the development of decentralized AI, a move that has led the company to appoint Shan Shan Wong, its Chief Operating Officer, and Christian Laforte, Chief Technology Officer, as interim co-CEOs. Stability AI, supported by investors such as Lightspeed Venture Partners and Coatue Management, is actively looking for a new permanent CEO. The organization, recognized for its influential image generation tool 'Stable Diffusion,' has recently encountered challenges in fundraising, particularly in securing new investments at a projected valuation of $4 billion. Mostaque expressed pride in the company's accomplishments, including achieving hundreds of millions of downloads and developing leading models across various modalities. He emphasized his continued belief in Stability AI's mission and expressed confidence that the company is well-positioned to maintain its commitment to open and decentralized AI under its new leadership.

Google's latest AI innovation, VLOGGER, is pushing the boundaries of generative media by creating realistic video avatars from single photographs. This technology, capable of producing high-fidelity video clips with accurate facial expressions and body movements, could transform customer service with more relatable helpdesk agents and revolutionize online communication, education, and personalized virtual assistance. However, VLOGGER also raises significant concerns about the potential rise of advanced deepfakes, as it can animate avatars to mimic real human behavior and speech closely. The development team at Google, led by Enric Corona, has engineered VLOGGER to surpass current avatar technologies by providing comprehensive facial expressions, gestures, and even upper-body movements, setting a new standard for lifelike digital human representation.

The Financial Times has introduced Ask FT, a new generative AI chatbot, designed to provide its subscribers with curated answers based on the publication's extensive archive of past articles. This tool differentiates itself by using specific and credible sources for its responses, rather than relying on a broader array of internet-based information, which may be less reliable or subject to legal issues. However, initial tests reveal some inconsistencies, such as outdated political information, indicating room for improvement in its accuracy. Currently available to a select number of subscribers in the FT Professional tier, Ask FT is powered by Claude, a large language model developed by Anthropic, though the Financial Times maintains a model-agnostic approach, meaning they could switch to a different AI model if it better suits their needs. Ask FT can answer a variety of questions ranging from current events to historical inquiries, pulling from decades of Financial Times journalism to provide answers with relevant citations.

A recent study by EduBirdie focusing on Generation Z's perceptions of AI in the workplace reveals a complex relationship. While Gen Z workers, those born after 1996, are utilizing AI technology, such as ChatGPT, to assist with tasks at work, a notable 36% reported feeling guilty about leveraging AI for work-related activities. Additionally, there's a concern among one-third of these individuals about becoming overly reliant on AI, fearing it might stunt their critical thinking abilities, with 18% feeling it could impede their creativity. This ambivalence might stem from their greater awareness of AI's ethical and social implications, reflecting their upbringing in a rapidly evolving technological landscape. The study also highlights an educational gap, with 20% of Gen Z workers facing challenges in using AI effectively in their roles, and a small fraction even facing job loss due to improper use of ChatGPT at work. Despite these concerns, many see the potential benefits of AI, with nearly half of the respondents acknowledging an increase in creativity and 14% noting a rise in their earnings thanks to AI. The most common uses for AI among this group include research, idea generation, content creation, and improving job applications.

🛠️ AI tools updates

WhatsApp is set to revolutionize the way users interact with images by introducing new AI-powered photo editing features. This upcoming update, currently in development and spotted by WABetaInfo, will allow users to apply artistic effects, change backgrounds, and utilize a 'restyle' tool for transforming photo styles to give them a fresh, artistic look. An additional 'expand' feature will enable the adjustment of photo sizes to better fit user needs. Emphasized by a trio of stars, the symbol for AI integration, these features aim to enhance user experience by offering innovative, customizable tools for messaging and sharing, all without additional cost. This strategy is part of WhatsApp's ongoing effort to maintain user engagement and attract new users by continuously introducing new and exciting features. Despite some concerns about AI manipulation of images, the ease of improving picture quality is expected to encourage widespread adoption of these tools.

💵 Venture Capital updates

Saudi Arabia is in negotiations with American venture capital firm Andreessen Horowitz to establish a massive $40 billion fund dedicated to AI investments, marking a significant move towards diversifying the kingdom's revenue sources beyond oil. This initiative aligns with Saudi Arabia's Vision 2030 goals, driven by Crown Prince Mohammed bin Salman, aiming to modernize the economy and reduce its dependency on oil. The talks, still in preliminary stages, explore forming a partnership between Saudi Arabia's Public Investment Fund, valued at $925 billion, and Andreessen Horowitz, a key player in Silicon Valley with $35 billion in assets and investments in almost 100 AI-related startups. This collaboration could reshape the global AI landscape, aligning with co-founder Marc Andreessen's view on the importance of pursuing global AI dominance to avoid considerable threats.

🫡 Meme of the day

⭐️ Generative AI image of the day