- AI KATANA
- Posts
- AI tools are designing entirely new proteins that could transform medicine
AI tools are designing entirely new proteins that could transform medicine
Also: Bill Gates isn't too scared about AI
AI KATANA
July 12, 2023
Greetings,
In today's round-up, we explore the promising and multifaceted applications of AI, from designing novel proteins to transforming labor markets and chip designs. RFdiffusion, an AI tool, is enabling the creation of unique proteins, potentially revolutionizing numerous fields, from medical treatments to synthetic cell signaling. The OECD warns that generative AI technology might fundamentally alter the workplace, highlighting the need for a coordinated response to balance risks and benefits. In the tech sphere, Bill Gates offers an optimistic outlook on AI's future, despite acknowledging the challenges it poses. Meanwhile, Lisa Su, AMD's CEO, believes AI will dominate future chip designs, a notion supported by key industry figures.
In AI tool updates, we look at AIDDISON, an AI-powered drug discovery platform aiming to expedite the discovery of suitable molecules and democratize AI-assisted drug discovery. Venture capital updates spotlight Sapphire Ventures' plan to invest over $1 billion in AI enterprise startups and the successful $5M seed round raised by Vellum.ai, a startup focusing on prompt engineering for generative AI.
Slicing:
👩🏻⚕️ AI tools are designing entirely new proteins that could transform medicine
🤑 OECD says rich economies on cusp of AI ‘revolution’
🤓 Bill Gates isn't too scared about AI
🦾 AI Will Dominate Chip Design
AI tools, such as RFdiffusion, are driving a significant shift in protein design, enabling the creation of novel proteins that could potentially revolutionize medical treatments, vaccines, and biomaterials. Using AI inspired by image synthesis software, these tools facilitate the quick design of new proteins that can bind tightly to specific biomolecules. RFdiffusion, developed by a team at the University of Washington, can create proteins based on specific criteria, generating innovative solutions and expanding the boundaries of what was previously possible. These AI tools are contributing to a new industry centered around protein design, with potential applications ranging from drug delivery to synthetic cell signaling. Despite impressive progress, challenges remain in designing more complex proteins and improving the success rate of these designs.
The Paris-based Organisation for Economic Co-operation and Development (OECD) has warned its member nations, some of the world's richest economies, to brace themselves for a potential upheaval in their labour markets due to the mass adoption of generative AI technology. With benefits such as productivity gains and higher job satisfaction being counterbalanced by job losses and ethical challenges, the OECD predicts that this 'AI revolution' could fundamentally alter the workplace. Highly skilled, white-collar jobs in fields like finance, medicine, and law, comprising about 27% of employment across OECD economies, are predicted to be most at risk. Despite the predicted growth spurred by AI, issues such as data protection, privacy, bias, and discrimination are cited as considerable concerns. The OECD emphasises the need for coordinated action to ensure the risks do not outweigh the benefits.
Bill Gates, in a recent post on his blog GatesNotes, outlined his stance on AI risks, focusing more on immediate concerns over long-term existential ones. Describing AI as the "most transformative technology" of our lifetimes, he highlighted current threats in fields such as education, elections, and employment. However, he maintains optimism about managing these risks, citing historical adaptations to technological changes like calculators, word processing, and spreadsheets. Gates suggests fast but cautious actions, calling for the creation of a global body to regulate AI, akin to the International Atomic Energy Agency, and support for retraining programs to cushion job market transitions. Despite this, critics argue his propositions lack novelty or specificity. While Gates acknowledges that AI will present a “bumpy transition," he confidently asserts that disruption to people's lives and livelihoods can be mitigated.
AMD's CEO Lisa Su stated at the 2023 World Artificial Intelligence Conference that AI will significantly shape future chip design given the escalating complexity of modern processors. Already instrumental in AMD's chip design process, AI is utilized especially in the 'place and route' stage for optimizing performance and energy efficiency, and is also employed in verification suites to hasten bug detection and improve test coverage during the chip's development process. Key industry figures such as Nvidia CEO Jensen Huang and AMD's CTO Mark Papermaster also see AI as an ideal tool for chip development. Additionally, leading electronic design automation toolmakers like Ansys, Cadence, and Synopsys are providing AI-enabled software to clients, with Synopsys launching Synopsys.ai, the first end-to-end AI-driven solution for chip development.
🛠️ AI tools updates
AIDDISON, an AI-powered drug discovery platform, seeks to revolutionize the laborious and costly process of small molecule drug development. The platform harnesses AI and machine learning in conjunction with traditional computational chemistry techniques like computer-aided drug design (CADD) to expedite the discovery of suitable molecules. It allows researchers to virtually explore the vast chemical universe, rapidly identify promising drug-like molecules based on their predicted activity, and estimate the synthetic feasibility of these compounds. Trained on experimentally generated datasets from pharmaceutical R&D, AIDDISON aims to democratize AI-assisted drug discovery by making it accessible not only to major pharmaceutical companies but also to smaller research entities worldwide. The integrated platform aims to reduce the time and cost associated with drug discovery, thereby accelerating the development of effective therapies for various diseases.
💵 Venture Capital updates
Sapphire Ventures is set to invest over $1 billion in AI enterprise startups from its existing funds, which total $10 billion under management with $3 billion poised for deployment. The investment will be primarily direct, but some will go to early-stage AI-focused venture funds via its limited partner fund. Firm president Jai Das emphasized the game-changing nature of AI for software development and pledged the company's commitment to building the necessary infrastructure for AI firms' success. Since OpenAI's ChatGPT's debut, $40 billion has been invested in AI startups, nearly a quarter of all startup funding, reflecting the global enterprise desire for AI technology despite data privacy and safety concerns. With previous investments in AI startups such as Clari and DataRobot, Sapphire aims to support firms making AI accessible through effective data use.
Vellum.ai, a startup focusing on prompt engineering for generative AI, has successfully raised $5M in a seed round with participants including Rebel Fund, Eastlink Capital, Pioneer Fund, and Y Combinator. The company, which was founded earlier this year, is seeing strong growth with around 25-30% monthly increase in revenue and a current client base of 40 paying customers. The founders, formerly employees at Dover, realized a significant market demand for efficient tools to enhance generative AI prompting while working with GPT-3, leading to the formation of Vellum. The startup provides AI prompters with tools for side-by-side comparison of model outputs, company-specific data search for context, and essential utilities like testing and version control, enabling the wider use of large language models (LLMs) in businesses. Vellum foresees a growing demand for such tools due to increased acceptance of natural language inputs in AI models, which could potentially expand the AI market size and open up opportunities for non-engineers to be prompt engineers.
🫡 Meme of the day
⭐️ Generative AI image of the day
