Cognition launches Devin, a generative AI-powered coding engineer

Hi!

Today's newsletter delves into the latest advancements and discussions in the world of artificial intelligence. Cognition AI Inc. unveils Devin, an AI-powered coding engineer set to redefine the landscape of software development with its autonomous coding capabilities. In the medical field, Johns Hopkins Kimmel Cancer Center presents ARTEMIS, an AI method for early cancer detection using liquid biopsies, potentially revolutionizing cancer diagnostics. The personal and ethical dimensions of AI are explored through Taiwanese singer Tino Bao's creation of an AI version of his late daughter, stirring debate over the use of AI in processing grief and memory. Meanwhile, the pharmaceutical sector witnesses a milestone with the commencement of clinical trials for the world's first AI-developed drug for a severe lung disease, underscoring AI's growing impact on drug discovery. In regulatory news, the European Union passes the AI Act, the world's first comprehensive AI law, setting significant standards for AI development and application. The newsletter also covers Google DeepMind's new AI, SIMA, aimed at enhancing gaming experiences, and updates from the venture capital world, including significant funding rounds for AI-driven platforms Cureskin and Prescient AI, highlighting ongoing investment in AI solutions across various industries.

Sliced:

• 🆕 Cognition launches Devin, a generative AI-powered coding engineer

• 🏥 AI Detects Cancer via DNA Repeats in Liquid Biopsies

• 😔 Taiwanese singer creates AI-version of late daughter, to sing happy birthday song to wife

• 💊 World’s first AI-developed drug for deadly lung disease enters landmark clinical trials in China, US

• ⚖️ European Lawmakers Pass AI Act, World’s First Comprehensive AI Law

🆕 Cognition launches Devin, a generative AI-powered coding engineer

Cognition AI Inc. has introduced a groundbreaking AI-powered coding tool named Devin, set to revolutionize the coding industry by autonomously creating code for entire engineering tasks, including AI models, without human assistance. This innovative tool surpasses the capabilities of existing developer aids like GitHub's Copilot, providing a comprehensive solution for completing coding projects and data research independently. Devin operates by seeking out educational content online to learn and execute tasks, with the ability to debug problems while allowing user intervention. The tool, currently in private preview, showcases the ability to handle complex engineering tasks by planning and executing steps within a sandboxed environment, effectively making thousands of decisions. Its introduction reflects a significant shift towards automating software operations, with potential implications for the future of human coders and the 'learn to code' paradigm. Cognition has secured a $21 million Series A funding to expand and improve Devin, promising a new era of AI-driven automation in software development.

🏥 AI Detects Cancer via DNA Repeats in Liquid Biopsies

A groundbreaking study from Johns Hopkins Kimmel Cancer Center introduces ARTEMIS, a machine learning approach capable of predicting early-stage lung or liver cancers by analyzing repetitive DNA sequences in both cancerous tissues and cell-free DNA from liquid biopsies. This innovative method could revolutionize cancer detection and monitoring by identifying changes in DNA repeat elements, often overlooked as "dark matter" or "junk DNA," which play significant roles in cancer and other diseases. ARTEMIS's potential was demonstrated in detecting altered genetic sequences in tumors and the bloodstream, offering insights into the cancer genome and providing noninvasive biomarkers for early detection and disease monitoring. The research, published in Science Translational Medicine, showcases ARTEMIS's ability to distinguish cancerous from normal tissues with high accuracy and to identify the cancer's origin, marking a new frontier in cancer diagnosis and personalized treatment. This study lays the groundwork for future clinical trials and highlights the untapped potential of analyzing genome-wide repeat landscapes for improving cancer outcomes.

😔 Taiwanese singer creates AI-version of late daughter, to sing happy birthday song to wife

Taiwanese musician Tino Bao utilized AI to create a video of his late daughter, Bao Rong, who passed away in December 2021 after a battle with a rare blood disease, to convey birthday wishes and a song to his wife. The project, motivated by the family's grief and the subsequent communication breakdown between Bao and his wife, represents a poignant application of AI in personal life. Tino Bao, who is pursuing a doctorate in AI, harnessed this technology to synthesize his daughter’s image and voice, with the latter being a challenging task given Bao Rong’s inability to speak during her final days. The initiative not only brought a form of solace to the grieving family but also sparked widespread online discussion about the ethics and emotional impacts of using AI to recreate the presence of loved ones. The use of AI for such deeply personal means highlights the evolving relationship between technology and human emotion, underlining the potential for AI to serve as a tool for expressing love and memory while navigating the complexities of grief and ethical considerations.

💊 World’s first AI-developed drug for deadly lung disease enters landmark clinical trials in China, US

An experimental drug developed using AI targeting a severe lung disease has begun phase 2 clinical trials in China and the United States, marking a global first for an AI-generated medication. This development, spearheaded by Insilico Medicine, demonstrates the significant impact AI can have on expediting and enhancing drug discovery processes. The company, which operates globally, including in Hong Kong, mainland China, Europe, the Middle East, and North America, asserts that the integration of AI has revolutionized drug discovery, making it quicker and more efficient. Insilico's CEO, Alex Zhavoronkov, emphasized the transformative potential of generative AI technologies in the pharmaceutical industry, highlighting years of research in applying AI to biomedical studies. This advancement not only signifies a milestone in drug development but also suggests a future where AI, robotics, and research into aging could intersect to discover treatments for complex diseases like Alzheimer's and Parkinson's.

🏛️ European Lawmakers Pass AI Act, World’s First Comprehensive AI Law

European lawmakers recently passed the world’s most comprehensive legislation on AI, marking a significant step in regulating the development and application of AI technologies. The AI Act, receiving final approval from the European Parliament, sets the stage for a safer, more human-centric approach to AI by introducing new rules for AI systems, including bans on certain uses, transparency requirements, and mandatory risk assessments for high-risk AI systems. With fines up to 7% of worldwide revenue for non-compliance, the law addresses both the risks and opportunities AI presents. Although it specifically applies to the EU, its effects are anticipated to be global, as companies operating internationally are likely to align with these regulations to maintain access to the European market. The legislation, which will phase in between now and 2027, also targets general-purpose AI models, requiring them to undergo safety evaluations and ensure data quality. This move reflects the EU's role as a pivotal player in setting global standards for technology governance, amid broader international efforts to address the complex challenges and potentials of AI.

🛠️ AI tools updates

Google’s new AI will play video games with you — but not to win

Google DeepMind has introduced SIMA (Scalable Instructable Multiworld Agent), an AI designed to play video games more like a human companion than a traditional, competitive AI. Unlike typical game AIs focused on winning, SIMA is trained to participate and follow instructions, adding a collaborative dimension to gaming experiences. It's currently being developed with skills for open-world and non-linear games, such as Valheim and No Man’s Sky, using natural language and image recognition to interact within these digital environments. By partnering with various game developers, Google aims for SIMA to adapt to different gaming styles without needing specialized APIs or game modifications. This AI endeavors to understand human-like gameplay through a blend of environmental interaction and response to verbal commands, demonstrating potential for more immersive and interactive gaming experiences. Although still in the research phase, SIMA represents a step towards integrating AI agents as dynamic participants in gaming, transforming how players engage with and experience video games.

💵 Venture Capital updates

Indian AI-driven dermatology platform Cureskin raises $20m from HealthQuad, others

Cureskin, an Indian AI-driven dermatology platform specializing in skin and hair health solutions, has secured $20 million in Series B funding led by HealthQuad, a healthcare-focused venture capital fund in India. The round also saw participation from existing investors such as JSW Ventures, Khosla Ventures, and Sharrp Ventures, bringing Cureskin's total funding to $26 million. This new investment aims to boost the platform's AI capabilities, accelerate growth, and broaden its solution offerings, according to CEO and co-founder Guna Kakulapati. Cureskin's innovative app uses AI for photo-analysis to assess skin conditions and their severity, addressing the shortage of qualified dermatologists, especially in smaller towns, and the common issue of ineffective self-treatments due to lack of product knowledge. HealthQuad, which has over $280 million in assets under management and is backed by Quadria Capital and KOIS, focuses on enhancing healthcare access and affordability through technology. The venture capital is also in the process of raising its third fund, targeting a corpus of $300 million, after closing its second fund at $162 million in 2022.

Prescient AI raises $10M Series A to help eCommerce brands optimize their ad spend

Prescient AI, an ad campaign revenue optimization startup, has successfully secured a $10 million Series A funding round, boosting its total venture capital to $18 million. The round was led by Headline and Delray Beach-based CEAS Investments, with contributions from existing investors Blumberg Capital, Focal VC, and several angel investors. Founded by Michael True and Cody Greco, Prescient AI aims to revolutionize marketing strategies for eCommerce and direct-to-consumer brands through its artificial intelligence-powered platform, promising more effective ad spend. The platform's unique selling points include rapid onboarding, precise cross-channel recommendations, and the ability to analyze initial datasets quickly. Notable for increasing ad spend return and driving revenue growth for brands like Khloe Kardashian’s Good American and online cookware brand HexClad, Prescient AI plans to use the Series A funding to nearly triple its team size and enhance its innovative marketing tools. This expansion aligns with the company's positioning at the crossroads of AI's rise and changing data privacy laws, aiming to become a leading solution for online advertising challenges.

🫡 Meme of the day

⭐️ Generative AI image of the day

Before you go, check out Singapore’s Ambitious AI Investment Plan.