Trump Chooses J.D. Vance as VP Pick, Praised for Open Source AI Stance

Welcome to Towards AGI, your premier newsletter dedicated to the world of Artificial Intelligence. Our mission is to guide you through the evolving realm of AI with a specific focus on Generative AI. Each issue is designed to enrich your understanding and spark your curiosity about the advancements and challenges shaping the future of AI.

Whether you're deeply embedded in the AI industry or just beginning to explore its vast potential, "Towards AGI" is crafted to provide you with comprehensive insights and discussions on the most pertinent topics. From groundbreaking research to ethical considerations, our newsletter is here to keep you at the forefront of AI innovation. Join our community of AI professionals, hobbyists, and academics as we pursue the ambitious path toward Artificial General Intelligence. Let’s embark on this journey together, exploring the rich landscape of AI through expert analysis, exclusive content, and engaging discussions.

TheGen.AI News

Trump Chooses J.D. Vance as VP Pick, Praised for Open Source AI Stance

Taking to his social network Truth Social, just days after narrowly surviving an assassination attempt by a gunman, former President Donald Trump today announced his running mate and vice presidential candidate for the 2024 U.S. presidential election: J.D. Vance.

Vance, a first-term Republican Senator from Ohio, is best known for his 2016 memoir, Hillbilly Elegy, which profiles the rural Appalachian region. He is also a former venture capitalist who received backing from influential contrarian Peter Thiel.

Vance’s views, which many, including the Democratic-voting writer, consider controversial and anti-democratic, include a prior campaign position to “end abortion.” However, he has garnered praise from the tech community for recent comments seemingly in support of open source AI and against regulation of the sector.

Last week, Vance testified during a U.S. Senate Committee on Commerce, Science, and Transportation hearing on “The Need to Protect Americans’ Privacy and the AI Accelerant.” According to TechPolicy.Press, Vance said:

“Very often CEOs, especially of larger technology companies that already have advantageous positions in AI, talk about the safety dangers of this new technology and urge Congress to regulate quickly. I worry that acting under pressure from these incumbents will benefit them, not the American consumer.”

Several tech figures on X interpreted Vance’s remarks as support for the development of open source AI without heavy regulation.

For instance, Brian Chau, executive director of the pro-open source nonprofit Alliance for the Future, posted one of Vance’s quotes on X. Self-described effective accelerationist Tetsuo summarized Vance’s remarks, stating, “The solution is open source.” Bindu Reddy, CEO of open source-based AI model provider Abacus AI, expressed her enthusiasm, saying, “Vance NAILS IT!” and added, “the solution, of course, is open-source!”

Software Developers Worldwide Ready for GenAI Revolution

A new study has found that an overwhelming 85% of the global software workforce anticipates using GenAI tools within the next two years. Conducted by Capgemini, the survey of almost 2,200 senior executives and software professionals highlights the extensive adoption of AI among developers, with the technology expected to greatly enhance efficiency.

Four out of five software professionals believe GenAI will significantly change their roles by automating simple, repetitive, and administrative tasks, allowing them to focus more on meaningful work. Although most organizations are still in the early stages of adoption, with 90% yet to scale their efforts, those already using GenAI have reported improvements such as increased innovation (61%) and higher software quality (49%).

The study also addresses ongoing concerns about the potential displacement of human workers by technology. Capgemini found that organizations plan to use the extra time gained from automation to develop new software features (50%) and upskill their workforce (47%), emphasizing that GenAI is intended to support human workers rather than replace them.

Reducing headcount is the least favored approach, with only 4% of organizations considering it. However, there are concerns about the security and legal risks associated with using generative AI in the workplace. Nearly half (46%) of software engineers are using GenAI, but two-thirds (63%) of these are using unauthorized versions, which could result in issues such as hallucinations, data leaks, and IP problems. Pierre-Yves Glever, Head of Global Cloud & Custom Applications at Capgemini, stated that while generative AI’s impact on coding efficiency and quality is evident, the true value will come from a holistic approach to software engineering. This includes addressing business needs with robust design, creating comprehensive developer workspaces and assistants, implementing quality and security measures, and establishing effective software teams.

How GenAI is Bringing Humanity Back to Healthcare at Kaiser Permanente

The potential for AI in healthcare is immense and impactful, but medical facilities should not rush its implementation, according to Daniel Yang, VP of AI and emerging technologies at Kaiser Permanente. Speaking at this year's VentureBeat Transform, Yang, who is also a practicing physician, emphasized the importance of a cautious, methodical approach to integrating this innovative technology.

Yang highlighted Kaiser Permanente's meticulous adoption strategy, noting their extensive clinical trial conducted across 21 hospitals in Northern California. The findings, published in The New England Journal of Medicine, were remarkable: the program was saving over 500 lives annually.

However, Yang pointed out that the algorithm alone did not save any lives—it was the redesign of the workflow that made the difference. “We had to design the entire workflow around it,” he explained. “That is 90% of the work.”

His team decided against alerting physicians directly, as it could be distracting and lead to false positives, wasting critical time. Instead, they trained critical care nurses to review patient charts and, if necessary, activate a rapid response team of doctors, nurses, and pharmacists to evaluate and determine the appropriate next steps.

When asked about generative AI, Yang humorously remarked, “I’m probably the first person to be on this stage that hasn’t mentioned gen AI in the first 30 seconds.” However, he noted that Kaiser Permanente is starting to implement this technology, particularly with a clinical AI scribe, which he claimed is the first of its kind in the country, “if not in the entire world.”

Yang highlighted a concerning statistic that burnout rates in healthcare range from 40 to 70%. With this in mind, he emphasized that one of the most exciting applications of generative AI in healthcare is not in treatment, diagnosis, or personalized medicine, but in reducing the administrative burdens on clinicians.

As a part-time practicing physician in internal medicine at Kaiser Permanente's urgent care clinic in San Francisco, Yang explained that during an eight-hour shift, he is expected to see 18 patients, which can feel like being on a “hamster wheel.” This constant demand can be extremely stressful, making the potential for AI to alleviate administrative tasks particularly significant.

Hong Kong Unveils New GenAI Tool as OpenAI Access is Restricted

The Hong Kong government is developing an AI tool similar to ChatGPT, initially targeting government employees but with plans for public use. This initiative is a response to OpenAI's decision to block access from Hong Kong and other unsupported regions.

Sun Dong, Secretary for Innovation, Technology, and Industry, announced the project on a radio show. The AI tool, called the “document assistance application for civil servants” in Chinese, is currently in testing and expected to be implemented across the government by the end of the year. The development is led by a generative AI research and development center at the Hong Kong University of Science and Technology, in collaboration with other universities. Future updates may include graphic and video design functions, although it is unclear how these will compare to ChatGPT. Sun emphasized the necessity of local development due to the lack of support from major tech companies like Microsoft and Google.

This AI tool development is set against the backdrop of the technological competition between China and the United States, both vying for AI leadership. China aims to be the global AI leader by 2030. However, OpenAI has excluded China, Hong Kong, and Macao from its supported regions, prompting the need for local solutions. OpenAI's recent restrictions on connections from these regions have not been fully explained. Local IT expert Francis Fong expressed cautious optimism about Hong Kong's AI capabilities, suggesting that while it may not rival ChatGPT, it could still achieve high standards with local input and customized features.

A locally developed AI system has the advantage of better addressing regional languages and issues. However, given the political climate, there is an expectation that such AI might adhere to certain ideological constraints, particularly in China, where internet content is heavily censored, and AI models must operate within these limits.

TheOpensource.AI News

Bioptimus Introduces H-optimus-0: The Largest Open-Source AI Model for Pathology

French startup Bioptimus has unveiled H-optimus-0, the world’s largest open-source AI foundation model for pathology. With 1.1 billion parameters, this model is trained on a proprietary dataset comprising several hundred million images from over 500,000 histopathology slides across 4,000 clinical practices. This sets a new benchmark in performance for various critical medical diagnostic tasks, such as identifying cancerous cells and detecting genetic abnormalities in tumors.

Rodolphe Jenatton, CTO of Bioptimus, stated, “Having launched less than five months ago, we are thrilled to introduce H-optimus-0, a fully open-source model that represents a significant advancement in pathology. We are dedicated to advancing medical research and improving patient care. By fostering collaboration and the adoption of new practices, we aim to accelerate innovations in pathology and beyond, ultimately benefiting patients worldwide.”

Pathology, the cornerstone of disease diagnosis, involves meticulous examination of tissue samples to identify abnormalities. Traditionally, this process relies heavily on the expertise and experience of pathologists. However, the increasing complexity and volume of cases necessitate advanced tools that can assist pathologists in making faster, more accurate diagnoses.

H-optimus-0 Key Features and Capabilities:

• Unmatched Scale and Performance: With 1.1 billion parameters, H-optimus-0 is the largest open-source AI model specifically designed for pathology, ensuring comprehensive analysis and high accuracy.

• Extensive Training Dataset: Trained on a vast dataset of over 500,000 pathology slides, H-optimus-0 has encountered a diverse array of cases, enabling effective generalization across different diagnostic scenarios.

• State-of-the-Art Diagnostics: H-optimus-0 delivers state-of-the-art performance in several key diagnostic tasks, often meeting or exceeding the performance of existing models. It was assessed on 5 tile-level tasks to identify tissue types or characteristics and 6 slide-level tasks to detect biomarkers or metastasis across various cancer types.

• Open Source Availability: As an open-source model, H-optimus-0 can be used by researchers to accelerate the development of new digital pathology models, fostering collaboration among researchers, clinicians, and developers, driving further advancements and solutions in pathology AI.

With its unprecedented scale and depth of training within the open-source community, H-optimus-0 utilizes cutting-edge AI technology to enhance the precision and efficiency of pathology diagnostics.

Professor Jean-Philippe Vert, PhD, co-founder and CEO of Bioptimus, remarked, “H-optimus-0 is just the beginning. It marks the first in a series of models we will develop at Bioptimus, each more advanced and comprehensive than the last. Future models will be trained on an even larger number of pathology images from Europe, Asia, and Africa and will incorporate other modalities, such as genomics and proteomics. Our ultimate goal is to create the first multiscale foundation model of biology, capable of integrating diverse biological data and scales to enable scientific discoveries and accelerate biomedical innovations.”

Closed AI

OpenAI’s Employee NDAs Under Scrutiny by US Regulators Following Whistleblower Complaint

Whistleblowers from OpenAI have requested the United States Securities and Exchange Commission (SEC) to investigate the company’s employee contracts, according to an AP report.

In a letter to SEC Chair Gary Gensler, the whistleblowers, remaining anonymous and confidential, questioned whether OpenAI has unlawfully restricted employees with non-disclosure agreements (NDAs) from discussing the risks associated with its AI technology. Sent on July 1, the letter urged the SEC to “swiftly and aggressively” enforce rules against NDAs that prevent employees from raising concerns with regulatory agencies.

The letter also referenced a formal whistleblower complaint recently filed with the SEC, initially reported by The Washington Post. AP obtained a copy of the letter from the office of US Republican Senator Chuck Grassley from Iowa, who confirmed that it was sent by legally protected whistleblowers.

Senator Grassley stated, “OpenAI’s policies and practices seem to inhibit whistleblowers’ rights to speak up and receive compensation for their protected disclosures. For the federal government to stay ahead of artificial intelligence, OpenAI’s NDAs must be revised.” OpenAI responded by stating that it supports employees' rights to make disclosures as per the law and has begun removing terms from contracts that imply punishment for “disparagement” after employees leave the company.

In May, OpenAI established a Safety and Security Committee, led by board members including CEO Sam Altman, as it begins training its next AI model, Reuters reported.

The SEC did not respond to inquiries, according to the AP. Microsoft has given up its board observer seat at OpenAI, citing significant governance improvements over the past eight months, as reported by Reuters. Apple, which announced the integration of OpenAI's ChatGPT into its devices last month, was expected to take on an observer role at OpenAI but has decided against it, according to the Financial Times, which cited a source with direct knowledge of the situation. Apple did not respond to a request for comment.

Project Strawberry: OpenAI’s Covert Effort to Revolutionize AI Research

OpenAI, the company behind ChatGPT, is working on a secretive AI project code-named "Strawberry," which aims to significantly enhance AI reasoning capabilities, potentially leading to breakthroughs in autonomous research and problem-solving.

Although details are tightly controlled, Reuters has obtained information from sources and internal documents revealing that Strawberry involves a novel method for training and processing AI models, enabling them to perform tasks that have been beyond the reach of current systems. Unlike existing models that focus on generating text-based responses, Strawberry is designed to give AI the ability to "plan ahead" and autonomously navigate the internet to conduct what OpenAI calls "deep research." This would mark a substantial advancement, requiring a deeper grasp of context, logic, and multi-step problem-solving.

The quest for human-level AI reasoning is a primary goal in the industry, with companies like Google, Meta, and Microsoft all exploring various techniques. Experts believe that achieving this breakthrough could allow AI to drive scientific discoveries, develop complex software, and address challenges that currently need human intuition and planning.

While OpenAI has not publicly disclosed specifics about Strawberry, a company spokesperson told Reuters, "We want our AI models to see and understand the world more like we do. Continuous research into new AI capabilities is common practice in the industry, with a shared belief that these systems will improve in reasoning over time." Strawberry appears to be an evolution of an earlier OpenAI project known as Q*, which generated internal excitement for its advanced reasoning abilities. Those who witnessed Q* demos reported its capacity to solve complex math and science problems beyond what current commercially available AI can achieve.

The exact mechanisms remain confidential, but sources indicate that Strawberry involves a specialized form of "post-training" — refining AI models after their initial training on massive datasets. This post-training phase, which may include techniques like "fine-tuning" and self-generated training data, is crucial for enhancing the AI's reasoning abilities.

OpenAI Acquires Rockset to Strengthen Data Analytics

OpenAI has purchased Rockset, a company specializing in real-time analytics databases, for nearly $200 million. This acquisition has sparked varied reactions from influencers on the social media platform "X," as reported by GlobalData, a leading data and analytics company.

Smitarani Tripathy, Social Media Analyst at GlobalData, commented: "Influencers view OpenAI's acquisition of Rockset as a wise strategic move that enhances its AI models with real-time data processing capabilities. They believe this positions OpenAI as a leader in innovation, marking a transformative step with the potential to unlock numerous opportunities. Influencers are optimistic about the merger, seeing it as a major advancement that could revolutionize AI-driven solutions, significantly expanding OpenAI's capabilities and opportunities. Additionally, others view this as a positive development for Rockset, solidifying its market presence."

Some notable influencer opinions captured by GlobalData’s social media analytics platform include:

Sarabjeet Johal, Technology Analyst and GTM Strategist: "@OpenAI acquires @RocksetCloud! Congrats to @GreylockVC and Rockset teams on a significant exit! Great pairing! Rockset enables businesses to leverage increasingly large and complex data sources. I've geeked out a few times at their annual conference in the Bay Area! A note…"

Muji, Founder of Hypergrowth: "OpenAI acquired Rockset over the weekend. This adds a data processing and hybrid search layer (vector db plus text and geospatial search) into the AI. Rockset competes with $ESTC $CFLT and analytics dbs like ClickHouse and Imply/Druid."

Matt Turck, Managing Director at FirstMark: "OpenAI acquiring Rockset (for 9 figures in stock) points to consolidation and exit paths in data/AI. Just one problem - Potential acquirers: OpenAI, Anthropic, Databricks, Snowflake, maybe Microsoft, a few others. Startups that would like to be acquired:"

Learn AI in 5 Minutes a Day

AI Tool Report is one of the fastest-growing and most respected newsletters in the world, with over 550,000 readers from companies like OpenAI, Nvidia, Meta, Microsoft, and more.

Our research team spends hundreds of hours a week summarizing the latest news, and finding you the best opportunities to save time and earn more using AI.

Sign up with 1-Click

Exploring the Future: Generative AI – A Deep Dive into Safe and Secure Production Deployment

Towards AGI is excited to announce the second event in their series, "Exploring the Future: Generative AI." As artificial intelligence continues to progress, generative AI is at the cutting edge of innovation, offering transformative possibilities across various sectors. This event, sponsored by Derisk 360 and hosted by Freshminds, will explore the latest advancements, practical applications, and future trends of generative AI.

The central theme of the event will be "Deploying and using GenAI applications in production in a safe and secure environment." The event is scheduled for July 31, 2024, starting at 5:30 PM, and will take place at the Freshminds Offices, Kingsbourne House, 229-231 High Holborn, London WC1V 7DA.

Sign up for the event now!

In our quest to explore the dynamic and rapidly evolving field of Artificial Intelligence, this newsletter is your go-to source for the latest developments, breakthroughs, and discussions on Generative AI. Each edition brings you the most compelling news and insights from the forefront of Generative AI (GenAI), featuring cutting-edge research, transformative technologies, and the pioneering work of industry leaders.

Highlights from GenAI, OpenAI, and ClosedAI: Dive into the latest projects and innovations from the leading organizations behind some of the most advanced AI models in open-source, closed-sourced AI.

Stay Informed and Engaged: Whether you're a researcher, developer, entrepreneur, or enthusiast, "Towards AGI" aims to keep you informed and inspired. From technical deep-dives to ethical debates, our newsletter addresses the multifaceted aspects of AI development and its implications on society and industry.

Join us on this exciting journey as we navigate the complex landscape of artificial intelligence, moving steadily towards the realization of AGI. Stay tuned for exclusive interviews, expert opinions, and much more!

Keep reading