TransformX Conference: Operationalizing AI at Scale

Use this time to make yourself familiar with the platform and network.

Greg Brockman, President and Co-Founder of OpenAI, will join Alexandr Wang, CEO and Founder of Scale, to discuss the role of foundation models like GPT-3 and DALL·E 2 in research and in the enterprise. Foundation models make it possible to replace task-specific models with those that are generalized in nature and can be used for different tasks with minimal fine-tuning.

In January 2021, OpenAI introduced DALL·E, a text-to-image generation program. One year later, it introduced DALL·E 2, which generates more realistic, accurate, lower-latency images with four times greater resolution than its predecessor. At the same time, it released InstructGPT, a large language model (LLM) explicitly designed to follow instructions. InstructGPT makes it practical to leverage the OpenAI API to revise existing content, such as rewriting a paragraph of text or refactoring code.

Before creating OpenAI, Brockman was the CTO of Stripe, which he helped build from four to 250 employees. Watch this talk to learn how foundation models can help businesses benefit from applications that they can create more quickly than with past generations of AI tools.

Thomas Kurian, the CEO of Google Cloud, will join Alexandr Wang, CEO and Founder of Scale, to discuss how AI helps businesses across various industries and use cases. Google Cloud is well-known for developer adoption, helping machine learning teams to create production-grade machine learning models. With platforms like Vertex AI and TensorFlow, Google boasts the most popular machine learning platforms adopted by over 3 million developers globally. Google also has succeeded with the widespread adoption of machine learning capabilities in its consumer and business products, including Gmail smart replies and predictive search.

Kurian will advise how to best roll out machine learning capabilities to many customers and ensure they are widely adopted. He will also discuss that, with the advent of foundation models, now is the time for all industries to more broadly adopt AI or risk falling behind the competition. He will detail practical use cases for retail, logistics, manufacturing, and healthcare. Kurian and Wang will also discuss the future of machine learning and what it will take to get there.

Before Google, Kurian spent 22 years at Oracle; his nearly 30 years of experience have given him a deep knowledge of engineering, enterprise relationships, and leadership of large organizations. Throughout his career, he has demonstrated a unique capability to align the latest technological developments, including machine learning, with real business problems to provide practical solutions to customers.

Leading organizations across every industry also need to be technology leaders, and today that means leveraging AI. Organizations applying AI don't necessarily have AI expertise, but they are prepared to leverage AI as a strategic asset for their innovation and product strategies. Many of these companies are looking for established use cases to supercharge their products, operations, and business lines. This is where pre-trained models create an opportunity for organizations without the resources and expertise build their own AI.

Alexandr Wang, Scale’s CEO and Co-Founder, will outline how Scale’s product investments are helping all types of organizations and teams apply and create their own AI.

Wang will also share how the latest advancements in AI allow next-generation models to be more versatile and attuned to specific business needs that weren’t possible before. He will explain why foundation models are the most exciting advancement happening in AI, and how Scale can help advance foundation models, especially as it applies to tuning these models to make them useful for specific tasks and industries including insurance, defense, logistics and many others.

Wang founded Scale while a 19-year-old student to help companies build long-term AI strategies with the right data and infrastructure. Under his leadership, Scale has grown to be valued as a $7 billion company serving hundreds of customers across industries from finance to e-commerce to U.S. government agencies.

Austin Russell, CEO and Founder of Luminar, will join Alexandr Wang, CEO and Founder of Scale for a fireside chat. The two will discuss the parallel missions of their respective companies with the purpose of training better models on high-quality, labeled data collected from the best 3D sensors available for autonomous systems.

Wang will ask Russell about his journey from Stanford dropout to his company’s 2020 NASDAQ IPO, product development, and how he thinks about building the machine learning infrastructure to train better models on 3D point cloud data. Russell will explain why he believes high-quality 3D sensors are essential for safe autonomy, and why simple 2D imagery doesn’t suffice.

Russell was a 2013 Thiel Fellow, which allowed him to drop out of his undergraduate studies and focus on Luminar full-time. He also became the world’s youngest billionaire as a result of his company’s IPO.

Modern medicine has provided effective tools to treat some of humanity’s most significant and burdensome diseases. At the same time, it is becoming consistently more challenging and more expensive to develop new therapeutics. The drug development process involves multiple steps, each of which requires a complex and protracted experiment that often fails. Insitro CEO and Founder Daphne Koller believes that, for many of these phases, machine learning models can help predict the outcome of the experiments and that those models, while inevitably imperfect, can outperform predictions based on traditional heuristics.

In this keynote, Koller will discuss how Insitro is bringing together high-quality data from human cohorts, while also developing cutting-edge methods in high-throughput biology and chemistry that can produce massive amounts of in vitro data relevant to human disease and therapeutic interventions.

Koller will also cover how the data are then used to train machine learning models that make predictions about novel targets, coherent patient segments, and the clinical effect of molecules. Insitro’s ultimate goal is to develop a new approach to drug development that uses high-quality data and ML models to design novel, safe, and effective therapies that help more people faster, and at a lower cost.

Koller also co-founded Engageli and has served as the co-CEO and President of Coursera. She received the MacArthur Foundation Fellowship in 2004 and the ACM Prize in Computing in 2008, and was elected a fellow of the American Association for Artificial Intelligence in 2004, in addition to receiving many other honors and awards.

Thanks to Dall-E, Stable Diffusion, Alexa, Siri, Google Assistant, and plenty of others, AI technologies are now more visible to the public than ever. But even though there have been numerous AI advancements and AI-infused products over the past decade, many organizations are still looking to understand how AI fits into their path to innovation. Eric Schmidt is an accomplished technologist, entrepreneur, and philanthropist and is well known for leading Google in its early days, from 2001 to 2011, and helping it to scale its infrastructure, diversify its product offerings, and maintain a strong culture of innovation. Today, he runs Schmidt Futures—a philanthropic venture—and advises governments and societies around the world on how to build a brighter future through technology.

Gerrit De Vynck, AI and algorithms reporter for The Washington Post, will moderate a fireside chat between Schmidt and Scale Founder and CEO Alexandr Wang, where they will talk about today’s advances in AI compared to 10 years ago. Schmidt and Wang will share their thoughts on what the next generation of AI companies will look like, what the proliferation of AI technology will mean for big tech, startups, and government, and how to push AI forward responsibly.

AI researchers have given the world a revolutionary new platform in the form of large models trained on internet data. And the game has changed dramatically with the recent release of advanced open source models.

Some people think that the model is the product. It is not. It is an enabling technology that allows new products to be built. The breakthrough products will be AI-native, built on these models from day one, by entrepreneurs who understand both what the models can do, and what people actually want to use.

While many AI technologies are still in their infancy, several are starting to emerge as practical for real-world applications. Many new AI-native products have grown explosively, reaching millions of users and creating huge businesses practically overnight. This is just the beginning.

Nat Friedman was the CEO of GitHub from 2018 to 2022, led Microsoft’s $7.5 billion acquisition of GitHub, and founded two startups. Friedman is joined by Alexandr Wang, Founder and CEO of Scale, to explore this new AI-native future and what it will take to succeed.

Dr. Craig Martell recently transitioned from an industry job as Head of Machine Learning at Lyft to becoming the new Chief Digital and AI Officer at the U.S Chief Digital and Artificial Intelligence Office (CDAO). Drawn by the mission, Dr. Martell brings a wealth of AI and business experience that he hopes will help increase the pace of technology adoption.

Dr. Martell will join Alexandr Wang, CEO and Founder of Scale, to discuss his charter to help the DoD increase its speed and agility for developing and fielding advances in AI, data analytics, and machine-learning technology. Dr. Martell will discuss how he plans to align incentives to take advantage of existing private-sector technology while streamlining procurement processes.

He will cover how his team will organize the immense quantities of data the department processes to make it more AI-ready to improve the department's capabilities overall.

Dr. Martell will also identify the advanced AI capabilities he plans to implement across the CDAO and how some of the latest advancements in large models factor into his plans. Finally, he will describe how his group communicates complex technical topics to non-technical leaders in the government and how the industry can best support the government with AI initiatives.

Deputy CTO of Microsoft and former CEO of Wikipedia, Lila Tretikov, joins John Maeda, CTO of Everbridge, to discuss how AI impacts humanity and how to best leverage this technology to grow. They will discuss the evolution of art and artists' role in the age of AI. Lila will discuss how Large Language Models (LLMs) are redefining how we interact with machines today and how advances in this technology will enable non-developers to build applications and immersive experiences with simply a prompt. Lila and John will explore AR/VR/XR, practical applications for headsets like Microsoft's hololens, and how this technology will evolve in the future. Lila and John also draw on their vast experience to outline the most impactful business applications of machine learning technology today.

They also discuss how AI mirrors humanity and the importance of being intentional about what data we use to train our models to enhance ourselves best. We think of AI as learning from us. But in its growing capacity to extend beyond any person's memory, computing, or even historical lifetime, it enables us to explore the sum of all human knowledge and to further simulate the possibilities of our future.

Neda Cvijetic, Senior Vice President of Autonomous Driving at Stellantis, will join Russell Kaplan, Scale’s Director of Engineering, for a fireside chat. The two will discuss the role of data diversity in building safer autonomy, and Cvijetic will share highlights from her career building autonomous vehicles at Tesla, NVIDIA, and now Stellantis, the parent company of Dodge, Fiat, and Chrysler.

Cvijetic will explain the opportunities that arise while building an entirely new infrastructure for training models for autonomous vehicles at a large automotive OEM, without the hindrance of having to support legacy systems. She will also discuss how she plans to achieve Stellantis’s publicly shared goals around Level 3 autonomy in 2024. The Stellantis portfolio includes Jeep, meaning its systems will also handle off-road scenarios. Kaplan and Cvijetic will also cover core paradigms in autonomous vehicles, from mapping to deploying to a million-vehicle fleet, to large language models, with more than a few surprising real-world anecdotes.

Prior to Stellantis and NVIDIA, Cvijetic worked on autopilot and infotainment systems at Tesla, served on the adjunct faculty of Columbia University, and held senior research positions at NEC Labs America. She holds more than 20 U.S. patents.

Stable Diffusion disrupted the deep learning scene when it was released in August, advancing the field of text-to-image models because of its ability to generate photo-realistic images given any text input. And unlike other models, Stable Diffusion makes its source code available, further advancing the democratization of AI. Stable Diffusion was created by Stability AI, in collaboration with EleutherAI and LAION. Stability AI is on a mission to design and implement solutions using collective intelligence and augmented technology and has developer communities with over 20,000 members who are building AI for the future. In this fireside chat, Emad Mostaque, CEO of Stability AI, will discuss emerging trends for open source AI infrastructure, the importance of data for real-world applications of AI, and predictions on the development of “text-to-everything” in artificial intelligence.

Mike Schroepfer, Meta’s first Senior Fellow, has been a driving force behind transformative technologies, from his time leading groundbreaking work in computer vision, natural language processing, and the metaverse to his focus today on combating climate change with technology. Schroepfer will sit down with Scale Founder and CEO Alexandr Wang to share lessons from Facebook’s transition from being web-focused to becoming mobile-focused and, ultimately, an AI-centric company. He will explain the developer pain that motivated an industry-standard framework like React Native, as well as the motivations behind keeping PyTorch as a community-focused tool, built for researchers.

He will discuss large language models and how more carefully curated, differentiated, and high-quality data will drive significant advancements in capabilities in the near term. He will also cover the future of augmented and virtual reality, including real-time machine translation with closed-captioning, contextual recognition of real-world objects to create more helpful applications, and more efficient hardware with native AI.

At Meta, Schroepfer focuses on supporting the company’s strategic technology priorities including its investments in AI and development of technical talent. From 2013 to 2022, he served as Meta’s Chief Technology Officer, where he led the development of the technology and teams that enabled the company to scale to billions of people around the world and make breakthroughs in fields like AI and virtual reality.

Pieter Abbeel wears many hats: Professor at UC Berkeley, Director of the Berkeley Robot Learning Lab, Founder of three companies, podcast host, and investor. The common thread is that Professor Abbeel is passionate about AI and robotics. In this keynote presentation, he will explore the possibility of training a large neural network to enable faster learning in robotics. Professor Abbeel will discuss his lab’s approach to solving this problem and will cover how video prediction is an excellent proxy for generalizable robots, the relevant models and datasets useful for pre-training, how unsupervised learning can help robots learn from themselves; and the usefulness of a human-in-the-loop. He will describe a four-step framework that might be able to lead, ultimately, to generalized robotics. Professor Abbeel is co-director of the Berkeley Artificial Intelligence (BAIR) Lab and founded Gradescope, which provides AI to help instructors with grading homework and exams, and Covariant, which provides AI for robotic automation of warehouses and factories. He is also a founding partner at AIX Ventures, a venture capital firm focused on AI start-ups, and is the host of The Robot Brains podcast, which explores what AI and robotics can do today and where they are headed.

Admiral William H. McRaven, a retired Navy four-star admiral and former SEAL, will sit down with Scale founder and CEO Alexandr Wang to discuss why America must regain its dominance in technology innovation – and how, by acting decisively, it can do so. In recent years, some other countries have outspent the U.S., but an effective policy is about much more than budgetary considerations. Admiral McRaven will discuss leadership lessons that apply to technology strategy based upon his decades of experience in special operations, his combat during Desert Storm and the Iraq and Afghanistan wars, his organization and leadership of the team that captured Osama bin Laden, and his years of advising U.S. presidents George W. Bush and Barack Obama on defense issues.

The Admiral has also written several books that wound up on The New York Times’ best-seller list, including Sea Stories: My Life in Special Operations; The Hero Code: Lessons Learned from Lives Well Lived; and Make Your Bed: Little Things That Can Change Your Life…And Maybe the World. The last was originally a commencement speech that the admiral repurposed into a book. Admiral McRaven is on the boards of the Council on Foreign Relations (CFR), the National Football Foundation, the International Crisis Group, The Mission Continues, and ConocoPhillips.

Few organizations have had a greater number of AI breakthroughs than Koray Kavukcuoglu’s team at DeepMind, a subsidiary of Alphabet. DeepMind’s research has changed multiple industries—from games with AlphaGo, to text-to-speech systems with WaveNet, and more recently biology with its AlphaFold protein-folding technology.

In July 2022, DeepMind announced that over 200 million predicted protein structures, representing virtually all known proteins, would be released on the AlphaFold database. AlphaFold can accurately predict 3D models of protein structures and has the potential to accelerate research in every field of biology.

Koray Kavukcuoglu is Vice President of Research at DeepMind. In this fireside chat with Alexandr Wang, Scale CEO, learn how his DeepMind researchers consistently bridge industry, science, and AI domains to create the world’s most meaningful AI advancements. Kavukcuoglu will discuss how AlphaFold evolved from research to having measurable real-world impact, the principles for managing incredibly large breakthrough projects and scopes, and how to plan for the next big scientific discovery.

Dr. Will Roper is a former U.S. Air Force and Space Force executive; he has spent his career implementing advanced technology—including machine learning and digital engineering, in both the government and private sectors. While he has seen first-hand that the Pentagon has made great strides in procuring advanced technology, Dr. Roper will describe the Pentagon not as digitally native but as preferring to say behind the innovation curve. He will suggest that the Pentagon needs to invest in digital infrastructure to unlock the most powerful machine learning capabilities on the battlefield. Dr. Roper will also lay out a roadmap for applying digital engineering and simulation to build advanced, battle-ready hardware more quickly.

As AI and machine learning advance rapidly, losing the innovation war will have drastic consequences for America's future. Dr. Roper calls for urgent investment in digital technologies to ensure that the United States stays ahead of other superpowers on the battlefield and beyond. Dr. Will Roper is currently a Distinguished Professor at Georgia Tech, focusing on technology that impacts national security. He previously served as Assistant Secretary of the Air Force, leading over $60 billion of annual technology development and operations for the U.S. Air Force and Space Force.

John List is a Professor at the University of Chicago and the Chief Economist at Walmart; he formerly served as Chief Economist at Lyft and Uber. Economists look at machine learning through a different lens. They identify causal effects, detect heterogeneity, and generally want to improve the bottom line and make the world better. List will take you through his experiences with machine learning and share the economics of apologies at Uber, the value of time at Lyft, and how the voltage effect impacts machine learning at scale. The Voltage Effect: How to Make Good Ideas Great and Great Ideas Scale is List’s latest best-selling book. To obtain data for the field experiments he pioneered in the 1990s, List has used several different markets, including charitable fundraising activities, the sports trading card industry, the ride-share industry, and the education sector, to highlight a few. This collective research has led to collaborative work with several schools, charities, and businesses, including Humana, United Airlines, Sears, General Motors, and many others. List’s research includes over 200 peer-reviewed journal articles and several published textbooks. He co-authored the international best seller, The Why Axis, in 2013. He is a current editor of The Journal of Political Economy.

Dr. Dan Patt, former DARPA Deputy Director and a strategist at the artificial intelligence company STR, will talk with Scale’s Head of Federal, Mark Valentine, about how the public and private sectors can better work together to maximize investments for national security. The scale, power, and efficiency of consumer products like mobile phones and their networks are improving defense capabilities, and the most exciting innovations are still to come.

Dr. Patt will explore several ideas like reducing the number of Department of Defense regulations to allow it to adopt enterprise AI faster, leveraging the “virtuous cycle” of consumer-oriented competition and product development, and funding “dual-use” development across both the commercial and defense sectors.

Dr. Patt has worked across the intersection of humanity and automation. As an aerospace engineer, he supported development of unmanned systems, including the A160 Hummingbird, and at DARPA led the development of attritable unmanned vehicles, and launched development of certifiable reduced-crew transport aircraft operations. As an entrepreneur, he was co-founder of Vecna Robotics, and is now Chief Strategy Officer at STR.

Bradley Horowitz, industry veteran and Google product leader, will talk with Scale founder and CEO Alexandr Wang about the transformative experiences that have helped him in his current role as lead for products with billions of users, which over the past decade have included Gmail, Calendar, Google Docs, Hangouts, Google News, Blogger, Reader, and Google Photos. He will discuss how he learned to build what users really need, and not what academics believe they need. He will discuss the potential of AI-enable features and applications being explored further and the need to solve the simple problems before being able to convince customers to try more sophisticated features.

When it comes to billion-user systems like Gmail, he will explain how he has helped tackle issues with data-driven product development, a good sense of triage, and relentless experimentation. He also will cover the lessons he learned from Flickr, which he helped Yahoo acquire, and how it’s best to let computers and humans do what they each do best.

Before joining Google, Horowitz helped lead video-analysis startup Virage to an IPO and served as VP of advanced development at Yahoo. Bradley is also an angel investor in Slack, Upstart, Grab, Miro, Coda, Sana and over 100 other high-impact startups.

Dr. Jane Pinelis, the Chief of AI Assurance at the Chief Digital and Artificial Intelligence Office (CDAO), will sit down with Scale's Head of Federal, Mark Valentine, to discuss the U.S. government’s efforts to develop explainable and ethical AI capabilities for national security. Dr. Pinelis will recount her journey to her current role, including the time she worked on Project Maven, which at the time was the Department of Defense’s most significant investment in AI, where she implemented a test and evaluation framework to evaluate vendor-submitted models. This work led to an AI assurance branch at the CDAO. Dr. Pinelas will discuss how the CDAO implemented AI principles of responsibility, equitability, traceability, reliability, and governability across the DoD and how it is working with international partners.

Dr. Pinelis will discuss the DoD’s plans for the next 12 months, including creating more best practices and tools to help implement advanced AI for different types of use cases including developers, senior leaders, and on the battlefield. Another focus is on the need to grow and nurture the existing AI workforce and build a strong base of responsible AI partners.

Dr. Pinelis will also discuss the launch of the Tradewind procurement ecosystem, allowing the DoD to accelerate the adoption of AI and machine learning by making it easier to work with academia, industry, and other groups.

Jeff Wilke is chairman and co-founder of Re-Build Manufacturing, a private company helping to bolster America’s industrial competitiveness. Before taking on this role, Wilke spent more than 21 years at Amazon, most recently as CEO of Worldwide Consumer, where he grew the business from $1 billion to a mind-boggling $386 billion a year. He spent his first seven years at Amazon building global operations. Wilke brings a wealth of knowledge across areas including supply chain, retail, and operations. Wilke will speak with Scale CEO Alexandr Wang to discuss his journey at Amazon and will cover how focusing on data and building models can help power eCommerce marketplaces. Learn how Wilke applied AI across the entire eCommerce stack and created an authoritative and comprehensive product catalog with more than 12 million listings. He will also share his predictions about the future of AI and sustainability. Wilke is an active philanthropist and mentor of startup founders who identify as female and/or People of Color. He is Vice Chairman of Code.org and Chairman of the U.S. Government’s Advisory Committee on Supply Chain Competitiveness.

Content comes in many forms—from traditional media with professionally developed content to social media with user-generated content, and everything in between. Over the past decade, media platforms have evolved significantly to keep up with the pace of newly developing content and content trends. For example, social media now support multiple forms of content, including images and short-form video, and traditional media platforms have expanded to support in-house content and streamed content. This expert panel will explore the importance of recommendation systems, discuss the importance of the data that powers these systems in the ever-evolving media platform experience, and cover how and why recommendation systems are so critical for growing media-based businesses.

The Department of Defense needs to become an attractive market for commercial companies and startups to apply innovative products and services in a meaningful way. Many potential technology providers believe the barriers to working with the DoD are too high. Congress and the department have created several initiatives and organizational structures to lower those barriers, but the DoD still struggles with transitioning innovation into production programs.

With so few examples of new technology companies gaining footholds in the DoD, many perceive that the department is engaged in “innovation theater” by pursuing the appearance of innovation while maintaining institutional resistance to the concept. Without a real culture of innovation, the DoD risks losing ground to other entities willing to leverage commercial technology.

This panel will discuss how maintaining political reputations and the fear of failure holds back innovation. They also will cover practical approaches for how the DoD can adopt pioneering technology more quickly, including revising the consortium model, providing better education about how AI and other advanced technologies work, and simplifying the procurement process to attract more private-sector technology companies.

The New York Life Insurance Company is the third-largest life insurance company in the United States, manages over $700 billion in assets, and invests in one of the most forward-looking data science teams in the insurance industry.

Glenn Hofmann is Chief Analytics Officer at New York Life and has spent over six years positioning New York Life's data science team to impact the bottom line, productivity, and even recruiting. Hofmann will join Melisa Tokmak, General Manager of Scale Document AI, to talk about New York Life’s data science journey. Learn how Hofmann has deployed advanced machine learning solutions that allow the company to easily access and analyze the data at scale.

New York Life’s data science team evaluates the risk taken on when issuing a life policy. The group works closely with the underwriting department to identify opportunities for AI algorithms to be used to help manage that risk. Among other achievements, the data science team helped build the infrastructure for data science models to be developed and deployed at a massive scale. Also, the team has created a data hub, a consolidated library of data assets, refined to an analyzable format.

Pinterest’s team knows that people want to feel included. When platforms lack representation, it tells people that the way someone may look or where they come from isn't the ‘norm.’ It’s more important than ever before to design inclusive systems that remove historical biases. Nadia Fawaz is the senior staff applied research scientist and technical lead of inclusive AI at Pinterest, which hosts over 400 million users who speak over 35 languages across 8 billion boards. In this keynote, Fawaz will explain how ML learns implicit bias and algorithmic fairness, how to design inclusive systems with cross-functional teams, how Pinterest learned from errors and made its models also learn from errors, and how Pinterest has built inclusive features, especially regarding skin tone and hair patterns, to create a more engaging platform. Fawaz will discuss how it is possible to change technology with intent. Fawaz’ research and engineering interests include machine learning for personalization, AI fairness and data privacy; her work aims at bridging theory and practice. Before joining Pinterest, she was a Staff Software Engineer in Machine Learning at LinkedIn, a Principal Research Scientist at Technicolor Research lab, and a Postdoctoral Researcher at the Massachusetts Institute of Technology’s Research Laboratory of Electronics.

Dr. Lynne Parker previously served as the Founding Director of the National Artificial Intelligence Initiative Office in the White House and coordinated AI policy across three different presidential administrations. She is currently the Associate Vice Chancellor at the University of Tennessee, Knoxville and Director of The AI Tennessee Initiative. She is joined in this fireside chat by Michael Kratsios, Formerly The 4th Chief Technology Officer of the United States and currently Managing Director at Scale.

Dr. Parker will explore concrete actions the federal government needs to take to accelerate the use of AI throughout the federal arena. These include better approaches to policy, investing in research and development, setting up programs for attracting AI talent to the government, and engaging across national and international borders. Dr. Parker also will talk about a risk-based methodology for defining better policy for the responsible use of AI, referencing the NIST National AI Initiative Act as a good model for this approach. Finally, she will lay out practical use cases where AI can make the government more efficient and effective, including reducing the burden of the vast amount of paperwork, combating malicious actors, and improving our geospatial intelligence and defensive capabilities.

She will also discuss how the AI Tennessee Initiative aims to position the state of Tennessee as a national and global leader in the data-intensive knowledge economy. Dr. Parker will address regional policy initiatives to ensure that AI benefits all Americans in areas such as jobs, healthcare, and education.

Mike Fisher, Chief Technology Officer of Etsy, will join Scale CEO Alexandr Wang for a fireside chat. Etsy’s systems process over 6 billion events each day. Etsy’s platform was originally created to link creators of unique items with people wanting to buy these one-of-a-kind pieces. In this fireside chat, Fisher will discuss scaling the Etsy platform to support over 100 million buyers and 5 million sellers using machine learning. Fisher also discusses how the company has rebuilt many of its search engine capabilities with more intelligent algorithms to provide more accurate recommendations.

Learn how Etsy addresses long-term product and content issues and how the company built a strong DevOps and engineering culture that encourages experimentation. Before joining Etsy, Fisher served as a captain and pilot in the U.S. Army, was a Vice President at PayPal, and had several technology leadership roles at GE.

While the spend management function has been around for decades, the industry is being rapidly disrupted by new, up-and-coming, customer-first vendors that distinguish themselves through technology. Spend management involves proactively and comprehensively supervising all supplier relationships and company purchasing to identify every dollar spent, and to get the most out of it. By listening to this panel, you will learn how leading FinTech companies are leveraging AI to not only transform their business operations but also to create a great customer experience. In this expert panel, you will hear from Henrique Dubugras, the Co-Founder and Co-CEO of Brex, and Dileep Thazhmon, the Founder and CEO of Jeeves, on how they view the spend management space. Dubugras and Thazhmon will discuss the market opportunity they saw with digital transformation and why they created Brex and Jeeves. They’ll talk about how they’re reinventing the space by creating a better customer experience with AI and share their insights on applying AI to the correct parts of each organization.

Learn how Dr. Jason Matheny, CEO of the RAND Coporation, and his team of researchers seek to make the world safer and more secure, healthier and more prosperous providing insights on advanced technology to policymakers. Dr. Matheny will sit down with Scale CEO and Co-Founder Alexandr Wang to discuss many of the urgent challenges facing AI, healthcare, and public policy today. They discuss advances in synthetic biology and AI, including DeepMind's AlphaFold, have an enormous upside potential for medicine, but also pose a threat because it makes this technology more available for bad actors.

Dr. Methany will also cover large language models and code generation tools, and how they will make developers and governments more efficient and more capable. He will also talk about whether AI’s offensive or defensive capabilities are more advantageous, and why public sector adoption of machine learning capabilities is so important. Other topics he will cover include how to ensure the US is a desirable destination for STEM talent including AI researchers, and how private sector technologists can provide value to policymakers to better understand technology and make more informed policy decisions. Dr. Matheny has served as Deputy Director of National Security, in other senior roles in the security field, and in various capacities in the healthcare industry.

Faire is an online marketplace that connects the world’s “best independent brands” with local retailers. There are over 85,000 brands and over 600,000 retailers using the platform across Europe and North America. Daniele Perito, Co-Founder and Chief Data Officer of Faire, will discuss how the marketplace is using machine learning to grow sales for both their small business customers and ultimately their consumers as well. Retailers on Faire can buy inventory and pay 60 days later, giving them the ability to return products that don’t sell, without disrupting their cash flow. Brands on Faire benefit from analytics and marketing tools that allow them to grow their customer base and their business. Offering these value propositions and acting as a matchmaker between brands and retailers requires dozens of models that span the gamut of machine learning and data techniques, from information retrieval to entity resolution to anti-fraud to default risk assessment with binary classifiers to operations research and econometrics.

Learn how Faire built and curated tools and infrastructure to be able to run thousands of trials simultaneously and experiment at scale. Prior to Faire, Perito was Director of Security, Risk for Square Cash, where he worked on building secure, fast, and easy-to-use products.

HSBC is one of the world’s largest banking and financial services organizations, with over 40 million customers across 60 countries. HSBC is in the midst of a transformational journey to become a digital-first bank. Learn how John Hinshaw, HSBC’s Group Chief Operating Officer, and his team have operationalized AI to create better customer experiences and make HSBC easy to do business with.

Two of the bank’s current AI applications include using AI for better global compliance with anti-money laundering regulations and reducing false positives in fraud detection processes, and another for predicting services customers might want. This followed HSBC introducing the world’s first AI-powered stock index family using IBM Watson in August 2019. Hinshaw will discuss how AI fits with HSBC’s digital and technology strategy, how HSBC has overcome the challenges of operationalizing AI at a massive scale, and the future of customer experience at financial institutions.

Before joining HSBC, Hinshaw served on the board of directors of the Bank of New York Mellon and chaired its technology committee. He also served as Executive Vice President of Hewlett Packard and Hewlett Packard Enterprise, where he managed technology and operations and was Chief Customer Officer. Julia de Boinville, Scale’s Head of UK Deployments, will host this fireside chat.

Sir Jeremy Fleming is the Director of the Government Communications Headquarters (GCHQ), the UK's intelligence, cyber and security agency. His mission is to keep the UK safe, and he is a strong advocate for collaboration between business and government that is supported by human-led technology. As technology continues to evolve, GCHQ is pioneering a new kind of security for an increasingly complex and interconnected world and is tackling a shifting and serious crime threat that is looking to exploit the pandemic.
In this fireside chat with Scale founder and CEO Alexandr Wang, Sir Fleming will discuss how he and his team apply AI to counter the world’s always-evolving cyber threats; balance privacy, regulations, and ethics as a government intelligence agency; his group’s open-source approach to AI; and the importance of helping humans guide AI development within rules of law that already exist. He will also cover the management lessons he’s learned that apply broadly, how technologists can use their skills to create a more peaceful world, and the ways AI is already being used to counter threats from terrorism to crime syndicates, After an early career in finance, Sir Fleming made his mark with distinguished service in MI5, modernizing the UK’s counter-terrorism strategy, and securing the 2012 London Olympics.

Interpreting pathology slides is a crucial part of diagnosing disease and an increasingly important part of drug discovery and drug development pipelines. The massive and ever-growing volume of data being produced for each of these areas has led to increased interest in harnessing machine learning to produce new biological insights, discover novel biomarkers, steer patient selection for clinical trials, and improve diagnostic accuracy. Effectively training and deploying computer vision models to interpret pathology images must overcome a massive hurdle: pathology slides are typically 100,000 x 100,000 pixels each, many orders of magnitude larger than is typical in computer vision pipelines. However, a number of approaches have been developed over the last few years that have made automatic interpretation of pathology slides not just feasible but a valuable tool used by many healthcare companies around the world. Nathan Silberman, ML Executive; Formerly, VP of AI at PathAI, Butterfly, will walk attendees through a number of these novel and creative approaches, including tiling. Prior to PathAI, Nathan was the VP of Artificial Intelligence at Butterfly Network and worked at Google Research, where he co-developed an internal library that became the Google TensorFlow 1.0 API.

During his 13-year tenure, Nick Beighton, former CEO of ASOS, grew the online apparel company from 200 million to 4 billion GBP in revenue by investing in data. Beighton will be joined by Jules de Boinville, Head of Scale’s UK deployments, to discuss how his team improved catalog data at ASOS and the key measurements they used to understand the business impact of their digital initiatives.

Beighton will also cover the core business values he helped create and best practices to effectively use customer and product data in eCommerce to deliver higher conversion rates. Beighton will describe how ASOS started by working on optimizing discovery tools, delivering product recommendations, and correctly tagging product attributes, four years before they invested in digital marketing. Underlying it all was a spirit of experimentation, which Beighton will explain in more depth. Before joining ASOS, Beighton worked for KPMG, specializing in business transformation, and held finance-related positions in other online retailers.

Want to take advantage of the latest research and advances in ML? Looking to chase the next big thing? Often the most useful machine learning models are based on the pain points you see every day. Tom Vu, Senior Director and Head of Data Science and Machine Learning at Flexport, has identified and applied disruptive machine learning opportunities that resulted in over $2 billion of value during the course of his career.

In this keynote, Vu will share his insights on how the limitations posed on human reasoning help identify ML opportunities, the importance of understanding processes and pain points relative to cognitive load, and what led him to decide to build an AI model to predict costs from CAD drawings. Vu’s portfolio of applied research projects and interests include routing, scheduling, assignment optimization under uncertainty, geospatial-temporal forecasting, imperfect information games, natural language processing, and computer vision. Prior to Flexport, Vu was the Head of Data Science and Analytics at WeWork, and the Chief Data Scientist at Boeing. He has over 20 years of experience implementing vision, transforming unmet business opportunities into realized software solutions.

Through the Patrick J. McGovern Foundation, Vilas Dhar is working to ensure that philanthropic causes get access to the same level of data literacy that already exists in the private sector. Dhar will sit down with Scale’s Head of UK Deployments, Jules de Boinville, to share an overview of the data-driven work his organization has performed around social justice causes and climate change. He will discuss joint research with the University of Chicago, which centers on examining economic outcomes as a function of education quality across the U.S.

Dhar will also cover the foundation’s efforts supporting Workforce Wanted, an effort to focus the group of 3.5 million data scientists on low- and middle-income countries, and machine learning coursework for Native Americans in the Dakotas who built a plant recognizer that incorporates their indigenous language and history. Dhar thinks back to exposing his grandfather in India to various modern technologies as an impetus to ensure that technology benefits everyone, not just the educated, wealthy, or privileged. Dhar has held many senior-level positions in several nonprofits seeking to build a more just world, including The Next Mile Project and the Berggruen Institute, and is on the board of directors of The Christensen Fund and the Network of International Donors.

AI is spurring a new wave of companies in construction tech that are transforming this $10 trillion-a-year industry. There’s substantial margin risk in construction when projects aren’t well planned, and this eats into overall economic growth, because construction represents 14% of the world’s GDP. The beginning of a project, known as “take-off,” is critical. Oleksandr Paraska, CTO at Togal.ai, will discuss how machine learning and computer science are improving efficiency in construction. Paraska talks about his journey in guiding his startup by integrating domain experts such as architects and engineering skill sets when building models. He also will explain how the nuances in architectural diagrams require object detection, classification, and segmentation to make sense of them. He also will share the challenges of scaling up processing volumes, in spite of the limits of GPUs, inference times, infrastructure cost, and MLOps resourcing. Paraska will describe some of the challenges in bringing machine learning to the construction industry, as well as the continuing need for custom code and why low ops/no ops tools just won’t work in this environment. In addition to his work at Togal.ai, Paraska is a freelance machine learning consultant at Tribe AI, and has worked as a machine learning engineer and developer at other companies.

The digital shelf has grown to include everything from hotels and vacation rentals to toothpaste, prescription drugs, corner shops, taco trucks, and even your neighbor's used couch. Over the past couple of years, eCommerce and retail companies have evolved the digital shelf and digitized commerce to meet consumers’ changing needs. Aatish Nayak, Head of Catalog at Scale, joins Saad Ahmed, Managing Director, Regional Head of Commercial at Grab, and Rong Yan, Chief Technology Officer at Verishop, to discuss the most pressing advancements. This expert panel will explore how platforms have evolved to digitize online retail, the importance of data-centric AI, and will share their predictions on the future of eCommerce.

Credo AI founder and CEO Navrina Singh will present two types of responsible AI: regulation-induced and value-induced. Regulations are changing rapidly—for financial services, there’s SR11-7, GDPR has AI implications, and regulation has become a top legislative agenda item. Some 60 countries are designing strategies and there are 1,000 responsible AI frameworks in development across OECD (the Organization for Economic Co-operation and Development) countries. Perhaps of most recent interest, Singh discusses New York City’s hiring bias law and what her customers are doing to comply by January 2023. Some regulations are still ambiguous, such as “disparate impact” referenced in the New York law. Singh explains how you can integrate appropriate bias monitoring and compliance tooling into your product to determine if you’re compliant and whether your latest model aligns with your business’s established values. Only then is it safe to deploy and serve your customers with a model. A technology leader with over 18 years of experience in enterprise SaaS, AI and mobile, Singh has held multiple product and business leadership roles at Microsoft and Qualcomm. Her company, Credo AI, specializes in trust, safety, and transparency for many industries—particularly for sensitive ones like financial services, human resources, and government.

Jordan Fisher, CEO of Standard AI, is passionate about changing the real-world retail experience by unlocking better shopping experiences with computer vision. Fisher will discuss applying computer vision models to the physical world to create human-centric applications; augmenting retail staff with better inventory tools and store layout analytics; and creating better shopper experiences with autonomous checkout. Even given the challenges in today’s retail environment—the labor shortage, inflation, the supply chain crunch, and difficulties competing against tech giants—shoppers expect more options and a better experience. Fisher’s vision is to allow customers to come in, shop, skip the checkout line, and get a receipt within minutes after they leave. At a shop at San Jose State University, traffic increased by 20%, average total receipts increased by almost 23%, and there was a decrease in wait time by over 50%. Further, store employees get analytics about which items are out of stock or misplaced, as well as traffic patterns about individual shoppers that do not collect PII. Fisher has spent his career focusing on both fundamental research and product development. He has worked in computational fluid dynamics, securities regulations, video games, machine learning, and retail, and seeks out areas where innovative products can be forged by tackling difficult research initiatives.

James Manyika, Senior Vice President for Technology and Society at Google, will sit down with Alexandr Wang, Scale founder and CEO, to discuss the best ways to ensure that AI models assist work and grow the economy, even as they replace certain mundane tasks. Manyika will describe how prioritizing utility and focusing on harm reduction should be at the core of principled use and deployment of AI. Manyika will point out the need to allow people to know whether they are interacting with an AI agent or a human, versus scenarios in which it’s permissible to use algorithms and human workers interchangeably. Learn how to adapt to new forms of work and activities that evolve as AI systems become more capable, how to develop AI policy to ensure no harm comes from the use of AI, and about the geopolitical impact and importance of AI principles. Manyika will also discuss his recent work for the Academy of Arts and Sciences, as editor of the Spring 2022 issue of Dædalus. For this journal, he invited a number of noted AI researchers and tech luminaries to write about different aspects of AI and society, including how the technology is being used for climate change, healthcare, and the economy—particularly how it can be made even better in those use cases.

Francois Chollet is a Software Engineer and AI Researcher at Google, created the Keras deep-learning library, and is a primary contributor to the TensorFlow machine learning framework. TensorFlow is the most popular machine learning platform, adopted by over 3 million developers globally. It is the third most downloaded repository on GitHub, runs on 4 billion devices, and powers machine learning at Google, Apple, Netflix, Twitter, and others. Keras is behind YouTube's recommendations and Waymo's self-driving cars.

Join Chollet as he shares what he considers to be the most important machine learning trends and their broader implications. Among the trends he will discuss is machine learning’s transition from an expert craft to a democratized utility, creating an ecosystem of reusable parts with premade and pre-trained models where all you need to do is bring your data. Another trend he will cover is the increasing scale of large models built on more data; this allows for use cases that are quickly changing our world. Simultaneously, the growing number of edge devices that can run ML models, including web browsers, is lowering costs, increasing privacy, and dramatically expanding the set of machine learning applications. Chollet will also talk about how, given all these other trends, it’s more important than ever to properly navigate privacy, bias, and safety issues to ensure beneficial applications.

Experts working on autonomous vehicles will discuss the rapid pace of research and innovation in machine learning, highlighting the most exciting developments over the last few years and how they approach incorporating the constant advances into an autonomous vehicle safely. The panelists are industry leaders working at NVIDIA, Tesla, Waymo, and Wayve. They will also discuss the unique approaches each OEM takes to leverage machine learning in their self-driving stack, with some using end-to-end learning and others preferring modular learning, and each method's advantages and disadvantages. They will also discuss best practices for integrating complicated sensor suites, software, data management, and machine learning with engineering. Other challenges they will cover include collecting large amounts of data, managing, and labeling datasets, integrating ML models with the rest of the self-driving stack, and how to improve the driver continuously. They will also discuss the critical role of simulation in development, and the current state of self-driving cars and the most difficult challenges they face today, including scaling to new environments quickly.

Something different and new is needed to accommodate AI’s full potential, and to monitor and regulate potential risks, especially given the geopolitical realities of today’s world. In this chat, political scientist Ian Bremmer—a best-selling author, the host of GZERO World, and president of the Eurasia Group—will expand on these views. Bremmer is perhaps best known for creating the J Curve, a popular theory that describes the relationship between a country's economic stability and its political openness. But the world is no longer defined only by nature and nurture, as Bremmer will explain; it’s also about algorithms. This is especially true regarding children—who they engage with and how they do so, how they think, and the groups to which they belong. Given this, and what Bremmer calls the absence of strong geopolitical leadership or the G-Zero, a new AI governance model is required. The world’s balance of power is no longer defined solely by the traditional heavyweights such as the US or China. Instead, the scenario developing now is the creation of multiple world orders with different key players in politics, security, technology and other areas. Two or three nations at a time will likely create treaties governing AI, Bremmer will explain, with input from the world’s largest tech companies, and those treaties will then grow to be adopted by other countries.

As the tempo of world events accelerates and the global threat landscape continuously evolves, defending our nation's interests is becoming increasingly challenging. In response, the DOD Digital Modernization Strategy seeks to advance digital innovation and embrace machine learning and artificial intelligence to ensure a competitive advantage over adversaries. Still, Federal entities face the same problem as the commercial world: operationalizing AI is hard. Unlocking the potential of machine learning requires dedicated ML Operations and advanced tooling to turn raw data into training data, rudimentary models into production-grade solutions, and research projects into mission-critical systems. For years, Scale has solved the most challenging machine learning problems and created cutting-edge tooling and innovative processes to support our customers. Now, we will showcase how Scale's ML Ops expertise enables state-of-the-art ground autonomy and overhead imagery analysis, enabling the DOD to quickly and efficiently get ahead of the global threat landscape.

Motional is making autonomous vehicles a reality; it was the first company to successfully perform a cross-country autonomous drive and has launched its publicly available ride-hail services in Las Vegas via the Lyft network. In this keynote, Laura Major, Motional’s Chief Technology Officer, will describe the various techniques the company uses to discover long-tail data—say, a cyclist carrying a surfboard or irregular driving patterns around construction sites—and incorporate that into both training and testing. Part of this involves combining sensor modalities including cameras, Lidar, and radar, Major will explain, and it all works with Motional’s Continuous Learning Framework. Major will give a bit of a preview of what’s coming next for Modality, including meal-delivery services and its plans for global expansion. Major began her career as a cognitive engineer for Draper Laboratory, where she combined her psychology and engineering skills to design decision-making support devices for US astronauts and soldiers. Laura also spent time at Aria Insights, Inc. (formerly known as CyPhy Works), a US-based drone manufacturer that specialized in developing highly advanced drones. She served as VP of Engineering, and then Chief Technology Officer, working on the development of autonomous aerial vehicles. She co-authored the book What To Expect When You're Expecting Robots: The Future of Human-Robot Collaboration.

Large Language Models (LLMs) have reshaped the AI landscape. They've shown incredible performance across copywriting, chatbot creation, text classification, code completion, and other tasks with the potential to impact every industry. It is increasingly necessary to leverage LLMs for efficiency gains and improved personalization, or risk falling behind the competition. Despite impressive progress in foundational models, it is clear that moving towards robust, safe, and reliable LLMs for production applications requires large amounts of high-quality labeled data for both training and robust evaluation. At Scale, we specialize in high-quality data collection, annotation, and reinforcement learning from human feedback (RLHF) to help you build or fine-tune models to stay ahead of the competition across LLMs, Computer Vision models, and more. In this workshop, Dean Shu, Head of Studio, and Shah Xia, Head of Rapid, will walk you through how to establish an efficient and effective data infrastructure to help you get your LLM production-ready, whether using Scale's workforce with Rapid or your in-house workforce with Studio.

Tim Ellis is the Co-Founder and Chief Executive Officer of Relativity Space and wants to put one million people on Mars. He plans on accomplishing this goal by using the 3D printed rockets manufactured by his company, which boasts a one-million-square-foot factory and the largest metal 3D printers in the world. Ellis will describe his first in-person rocket test and how it drove him to build Relativity Space. He will also outline the concept of software-defined manufacturing, and how combining 3D printing and machine learning allows Relativity Space to experiment and iterate more quickly than traditional manufacturing methods allow. He will discuss how Relativity collects data and how sensor fusion helps the company predict printing flaws and identify them when they happen. In this fireside chat with Scale CEO and Founder Alexandr Wang, Ellis will discuss how ambition can make absurd goals a reality and what it takes to persevere at a startup. Finally, Ellis will outline the benefits we will see along the way to achieving his goal. Ellis is the youngest member of the National Space Council Users Advisory Group, which advises the White House on space policy. He also serves on the World Economic Forum as a Technology Pioneer.

As CTO and Co-Founder of Embark Trucks, America's longest-running self-driving trucking program, Brandon Moak will discuss both the opportunities and the challenges involved with bringing a fully autonomous truck fleet to production. Many of today’s generation of professional truck drivers are over the age of 50, and global supply chain shortages are continuing, so this industry is ripe for automation.

Moak has helped Embark expand and mature its machine learning capabilities from rudimentary to sophisticated. But it's not been easy, as Moak will cover; the company at one point needed to take a step back and figure out the essential, repeatable, data-based elements of successful projects before it could scale up and out in a meaningful way. Moak will describe how Embark generates a new model from scratch, in this case a Lidar segmentation detector. You will learn how the company built a high-quality, fleet-wide data engine that ensures accuracy and consistency in data collection and annotation, how it deployed upgraded sensors across a large fleet, how it implemented advanced dataset curation tools to mine interesting data and conquer the long tail of edge cases, and how it manages and motivates an expanding machine learning workforce.

Before joining Embark, Moak previously held engineering and senior management positions at Kindred.ai and Clear Blue Technologies.

As the Director of AI Research at NVIDIA, Dr. Anandkumar will share some of the highest-impact topics her team is researching—from better weather modeling to tackle climate change, to modeling CO2 carbon capture and storage, to coronavirus aerosol simulation and drug discovery. Previously, Anandkumar helped launch Amazon SageMaker, Comprehend, and Rekognition at AWS (Amazon Web Services). She is also the Bren Professor at Caltech’s CMS (Computing + Mathematical Sciences) Department and serves as part of the expert network of the World Economic Forum.

Building on her seminal paper that covers the initial development of tensor algorithms, Dr. Anandkumar will present her research on Fourier Neural Operators (FNOs), which in some cases can replace computationally costly Navier-Stokes equations that underpin many fluid dynamics simulations such as weather forecasting models and drug discovery processes. She advocates the use of the Fourier transform in neural networks to make them discretization- (or even quantization-) invariant. By replacing brute-force computational approaches with FNOs running on GPUs, Dr. Anandkumar is able to reduce the complexity of the simulation of weather modeling (a 45,000x speed-up) and carbon capture and sequestration (a more than 10,000x speed-up).

Join Dr. Anandkumar as she’ll inspire attendees to apply FNOs, smarter model architectures, and parallel computation to solve the public health and climate crises of our time.

For industries that rely on paper documents, digital transformation delivers faster processing times and unlocks new workflows. While many companies are making sizable investments to transform their business, most are experiencing challenges trying to make outdated technology work. Intelligent document processing (IDP) powered by AI is proven to reduce total cost of ownership and increase efficiency across freight forwarding, trucking, expense management, lending, insurance, customer onboarding, and more. In this workshop, learn from AJ Ostrow, Head of Document AI, and Jessica Zhu, Lead Engineer of Document AI Go, about how to evaluate IDP and optical character recognition (OCR), best practices to handle complex and unstructured document types, and how to use Document AI to scale processing for large volumes and workflows.

While the current era of machine learning is focused on recognizing patterns, the future will be all about making decisions, Michael I. Jordan, a Distinguished Professor at UC Berkeley, will explain in this keynote. However, it's not easy to get there. Real-world decisions have consequences, are powered by context, are often interlaced with others’ decisions, have multiple levels of related decisions, and must be explainable. He will talk about how today’s recommendation engines, for instance, can wind up sending hundreds of people to the same restaurant or movie, or along the same traffic route. Professor Jordan will cover how the future of machine learning relies on economic principles. He will discuss how decision-making is different from recommendations, how market forces factor into decision making, and will explain a framework for building a decision-making system. Professor Jordan’s research interests include machine learning, optimization, control theory, and computational biology. He is a member of the National Academy of Sciences, the National Academy of Engineering, and the American Academy of Arts and Sciences. He has received the Ulf Grenander Prize from the American Mathematical Society, the IEEE John von Neumann Medal, the IJCAI Research Excellence Award, and the ACM/AAAI Allen Newell Award.

Autonomous trucking is a high-stakes industry. Truck crashes cost $30 billion annually, the U.S. has a shortage of 80,000 truck drivers, and approximately one-third of trucks drive so-called empty miles, or the mileage that accrues while driving an empty container or trailer and not earning revenue. Building a more capable driver is essential to safer roads and is one of today's most complex technical challenges.

Waymo Via's head of trucking, Boris Sofman, will share his insights about how the Waymo Via team is building a better autonomous truck driver, which includes maintaining a single-driver platform across passenger vehicles and commercial trucks. Sofman will also discuss how his company leverages synthetic data to improve the driver in rare scenarios, including different weather situations, construction scenes, and sensor degradation. He will also cover his team’s use of advanced sensor suites, data collection, and labeling techniques, including using multiple vehicles for better ground truth.

Prior to Waymo, Boris was the Co-Founder and CEO of Anki, where he and his team developed various AI-based entertainment products and shipped over 3.5 million robots and devices around the world. In other positions, he also worked on off-road autonomous vehicles and ways to leverage machine learning approaches to improve navigational capabilities in real time.

Collective intelligence studies the group brainpower that emerges from the interactions of many individuals. It's commonly observed in nature—for example, when a group of fish decides which direction to swim or when elephants choose where to migrate. Google Brain's David Ha, a research scientist, will share methods for using collective intelligence to improve today’s deep learning models.

The current generation of neural network models achieves state-of-the-art performance on tasks across fields spanning computer vision, natural language processing, and reinforcement learning. But as these models become larger and more complex, they suffer from issues including poor robustness, the inability to adapt to novel task settings, and other problems. Collective behavior, however, tends to produce systems that are robust, adaptable, and have less rigid assumptions about their environment configurations.

In this keynote, Ha will highlight several active areas in modern deep learning research that incorporate the principles of collective intelligence to advance current capabilities, including lessons in deep reinforcement, multi-agent, and meta learning. He will provide examples from Reddit, Conway's Game of Life, Minecraft, Puzzle Pong, and more. Ha works in the Google Brain team in Japan, where his research interests include complex systems, self-organization, and creative applications of machine learning. Prior to joining Google, Ha worked at Goldman Sachs as a Managing Director.

Machine learning leaders from robotics (Covariant), home automation (Resideo), autonomous delivery (Nuro), and warehouse automation (Pickle Robot) sit down with Russell Kaplan, Scale’s Director of Engineering, to share their approaches to dataset management. Pickle Robot CTO Ariana Eisenstein will share how she thinks about modulating quantities from different data sources like synthetic and public open datasets with real-world data for training datasets. Mostafa Rohaninejad, Founding Research Scientist at Covariant, will describe how the object “picking” problem requires synthetic data for unsafe scenarios and how he also incorporates structured and time-series data—supervised and unsupervised learning should go hand-in-hand. Jack Guo, Head of Perception at Nuro, will explain how it’s essential to have tools and mechanisms to automatically highlight recorded data that deviates from the norm, especially if it was captured in a new location. Like Rohaninejad, he will stress the importance of simulation as a component of successful reinforcement learning. Louis Tremblay, AI/ML Engineering Leader at Resideo, will explain how security cameras in the home represent an even more unbounded environment than do warehouses.

The group will also discuss why maintaining separate datasets and training pipelines for different customers is both costly and incurs additional technical debt over time. Testing on fault-tolerant customers first before deploying to the wider fleet is also important. Scale’s Kaplan will share how, in his experience, when metrics and anecdotes seem at odds, it makes sense to re-think the metrics and establish new ones.

The pace of user-generated content (UGC) is increasing exponentially. While social media and content platforms are innovating to meet consumer needs, the acceleration of development without AI is insufficient in creating robust recommendations and personalization systems. Furthermore, UGC creates a risk for content platforms when content violates platform guidelines. Content understanding is critical to uncover content trends early, tailor personalized recommendations, and increase trust and safety. In this workshop, learn from Lauren Oh, Product Manager of Content Understanding, about how to use data to deliver the most engaging user experience. Attendees will learn how to identify content growth areas to boost engagement across social media, eCommerce, and web3 marketplaces.

Over the past couple of years, the consumer robotic market has seen rapid adoption due to the pandemic and dropping prices, and the market is expected to grow at a CAGR of almost 31% through 2027. Last year, Amazon announced Astro, a new household robot for home monitoring that combined AI, computer vision, sensor technology, and voice and edge computing.

Astro brought new advances to the consumer robotic market around human-robot interaction and multimodal AI. In this fireside chat, Dr. Ken Washington, VP of Software Engineering for Consumer Robotics, will discuss scaling robotics at Amazon and new features in Astro that are novel in the areas of computer vision, perception training, and mapping. Dr. Washington will be joined by Vijay Karunamurthy, Head of Engineering at Scale AI, to discuss the future of robotics at Amazon and the new AI technologies evolving with consumer robotics.

Before joining Amazon, Dr. Washington was CTO at Ford Motor Company, where his portfolio included propulsion systems, sustainable and advanced materials, additive manufacturing, next-gen vehicle architectures, controls, and automated systems. Dr. Washington also served as Lockheed Martin Corporation’s first chief privacy officer and as CIO for Sandia National Laboratories. In addition to his election to the NAE in 2020, he received the 2012 Black Engineer of the Year Award in Research Leadership.

Eric Rachlin, Body Labs’ Co-founder, and Natalya Tatarchuk, Distinguished Technical Fellow and Chief Architect at Unity, will join Vivek Muppalla, Scale's Director of Synthetic Systems, to discuss the current and future state of digital humans, avatars, and AI. Rachlin and Tatarchuk will describe how to create more realistic digital humans with methods including traditional motion capture, motionless capture, and recent machine learning techniques like Instant NeRF. They will also discuss supplementing models with synthetic data for improved performance. They will outline how text prompting will evolve from state-of-the-art diffusion models to more powerful digital creation tools, allowing non-developers to create and animate realistic digital humans and to control and customize every element of digital environments to create more immersive experiences.

The panel will address Metaverse standards such as Universal Scene Description and other machine learning methods to provide an "intelligent glue" to help provide smooth cross-application interactivity and realistic object interactions for avatars. Rachlin and Tatarchuk will share various applications for digital humans, including virtual try-on, inclusive and diverse digital assistants, and more realistic digital avatars for gaming. They will address mitigating bias, distinguishing between real and fake identity, and protecting privacy. They will end with thoughtful advice for anyone who wants to get started in the field.

eCommerce and Retail companies are shifting to prioritize the digital shelf as US eCommerce sales will cross $1 trillion for the first time in 2022. While eCommerce companies are innovating to meet consumer demand, retailers are still contending with challenges in building robust search, personalization, and recommendation systems for online shopping. At the core of scaling marketplaces is starting with robust and expansive product catalog data. In this workshop, learn from Aatish Nayak, Head of Catalog, and Mahesh Murag, Product Manager, Catalog, about how high-quality catalog data and data-centric AI helps eCommerce teams build great customer experiences. Attendees will learn insights and best practices for maximizing and improving operational efficiency in retail through a live showcase of our retail product.

Richard Socher has been a transformative force in AI, from his academic publications to his work at MetaMind, Salesforce, and his current stint at you.com, a new search engine that combines privacy, freedom from ads, and ‘trust, kindness, facts, and AI.’ During this fireside chat with Russell Kaplan, Scale’s Director of Engineering, Socher will discuss how recent developments in AI are shaping our world and how to develop the next generation of AI talent.

Socher will talk about the latest developments in AI, including AI-powered code completion tools, large language models, and AI applications for healthcare. He also will address how to build large-scale ML teams and what it means to invest in the future of AI. In addition to his ongoing work as CEO of you.com, Socher is a key member of AIX Ventures, which helps AI-powered companies raise money from AI practitioners. While at Salesforce, Socher led teams working on fundamental research, applied research, product incubation, CRM search, customer service automation, and a cross-product AI platform for unstructured and structured data.

Join this enterprise-focused, spirited discussion on how best to train, use, and fine-tune foundation models in the enterprise. Elliot Branson, Director of Machine Learning & Engineering, Scale AI, will moderate the panel with industry experts from AWS, NVIDIA, Netflix, and Meta.

Erhan Bas, formerly Applied Scientist at Amazon Web Services and now at Scale, will share his perspective on training large language models (LLMs). Bryan Catanzaro, Vice President of Applied Deep Learning Research at NVIDIA, will share how the GPU manufacturer is targeting foundation models as a core workflow for enterprise customers. Faisal Siddiqi, Director of Machine Learning Platform at Netflix, will share how his company is using foundation models to analyze highly produced video content. Susan Zhang, Researcher at Facebook AI Research (FAIR), a division of Meta, will share insights from training and fine-tuning Meta’s OPT model.

Members of the panel will share how they scale their training across multiple nodes, attempt to avoid overfitting by mitigating data quality issues early on, and address bias in models trained on a large internet-based text corpus. The panelists will discuss the compute cost inherent in training an LLM from scratch, how to avoid costly and tedious hyperparameter optimization, the need to mitigate training failure risk in clusters with thousands of GPUs, including sticking to synchronous gradient descent, and the need for extremely fast storage devices to save and load training checkpoints.

By 2026, 25% of people will spend at least one hour a day in the Metaverse, according to Gartner. These users will spend their time in new AR/VR experiences at work, shopping, education, social media, entertainment, and fitness. However, building the Metaverse is hard: Collecting, labeling, and managing the large volumes of data needed to create truly immersive Augmented Reality (AR), Virtual Reality (VR), or Extended Reality (XR) experiences is expensive and challenging even for large companies. Scale enables our customers to develop sensational XR experiences more quickly and efficiently with our industry-leading data collection and labeling solutions. This session will demonstrate how to build an AR assistant by enabling the collection of large volumes of high-quality human-centric video footage; how to annotate large volumes of data with multimodal data annotations, including dynamic cuboids, dense point cloud segmentation, and human keypoint annotations; and synthetic data generation for mitigating bias and optimizing your data collection and labeling investments.

Sriram Raghavan, Vice President of IBM Research for AI, will discuss foundation models with Vijay Karunamurthy, Scale’s Director of Engineering. Learn how Foundation Models allow enterprises to leverage recent research, while also fine tuning to their respective use cases and (often smaller) training datasets. Raghavan will explain how the effectiveness of Foundation Models comes from their ability to generalize nearly any data type and domain as another language. This is the same flexibility that helped IBM Watson double its number of supported languages in a year. Since then, as Raghavan will cover, IBM has demonstrated that it’s surprisingly effective to extend foundation models to time series and tabular data, simply by treating them as new forms of language.

He will also discuss how to bring trust into the training process when working with transformers, and the most effective way to employ synthetic data for specific use cases. Prior to his current role, Raghavan was the Director of the IBM Research Lab in India and the CTO for IBM in India/South Asia. He began his career in IBM at the Almaden Research Center in San Jose, California where he led a number of research efforts at the intersection of natural language processing, data management, business analytics, and distributed systems.

Abnormal Security builds ML products that help protect systems against cyber attacks. Dan Schiebler, Head of Machine Learning at Abnormal Security, discusses best practices for building cybercrime detection algorithms. In this session, Schiebler will cover how to design, monitor, and launch resilient ML systems and how to train ML models on production issues. He will talk about the different types of problems that production ML systems can encounter, including features that become unavailable because of upstream data issues, distribution changes, or features that become stale. Schiebler will address the different types of iteration loops in most companies—online vs offline—and how that plays into testing and training, as well as the company’s ablity to tolerate risk. Historical logs and data also play a key role.

Before joining Abnormal, Schiebler worked at Twitter: first as an ML Researcher working on recommendation systems, and then as the Head of Web Ads Machine Learning. Before Twitter, he built smartphone sensor algorithms at TrueMotion.

Scale Product Manager Ben Levin and Product Engineers Tim Lu and Alan Yu will show how using aerial imagery-based maps in conjunction with overlapping lidar scenes helps to avoid alignment issues and calibration errors in training data. Using the example of autonomous navigation, the team will show how you can reduce your labeling spend while increasing data quality by aligning top-down annotations with 3D point cloud scenes. Ben Levin will highlight features in our software stack that ensure that mapping labels are accurate, even across large geospatial regions, in spite of varying geography. Tim Lu will show how to use aerial imagery to annotate 3D lidar point clouds more efficiently, and Alan Yu will explain how overlaying 2D imagery can help scale up the 3D lidar annotation process, since 3D point cloud scenes typically cover a smaller area than simpler and larger 2D images. Attendees will learn best practices from our experience designing hybrid labeling pipelines for a large number of automotive customers.

Humans evolved to communicate with other humans, not with algorithms. So, to create algorithms that are better at doing what we want, we need to understand how humans communicate. Alan Cowen, CEO and Chief Scientist at Hume AI, will discuss how algorithms can better understand human communication and the role this will play in the future. Cowen will cover semantic space theory, a new data-driven way of thinking about emotions and how we express them. Hume runs experiments all over the world to see how humans express themselves, and measures nuances of expression in voice, language, face, and body movements. Hume then leverages this data to fine-tune models, such as GPT-3 to create a model that controls the emotional tone of a response. The goal is to create more responsive assistive technology and to enhance training tools for healthcare professionals and others. He will also discuss what new datasets and models are teaching us about the vocabulary of human expression and some of Hume’s findings, and how expressive communication and empathy can be built into modern technology. Cowen is an applied mathematician and computational emotion scientist. Prior to founding Hume AI, he was a researcher at Google AI, where he helped establish affective computing research efforts.

Text embeddings are useful features in many applications including semantic search, predicting code completion, natural language, topic modeling, classification, and computing text similarity. Arvind Neelakantan, Research Lead and Manager at OpenAI, will introduce the concept of embeddings, a new terminus in the OpenAI API.

When OpenAI originally introduced the API two years ago, it was based on the GPT-3 model, which was useful for many tasks. But, as Neelakantan will explain, GPT-3 is not explicitly optimized to produce a single vector or embedding of the input. This ability, to have a condensed representation of the input, would be helpful for programmers and others to use as features for downstream applications, the OpenAI team determined. They set about building an unsupervised model that is good at getting this kind of single embedding, and created a contrastive pre-training model, which Neelakantan will describe. He will also cover use cases for embeddings, and how the API is used in the real world, including at JetBrains Research for astronomical research and at FineTune Learning, which builds education systems. FineTune is using text embeddings to more accurately find textbook content based on learning objectives.

The rise of NFTs have exploded over the past year. The NFT market is expected to grow by over $140 billion over the next four years. OpenSea is the world’s first and largest Web3 marketplace for NFTs and crypto collectibles. During this fireside chat, Shiva Rajaraman, Vice President of Product at OpenSea, will join Vijay Karunamurthy, Head of Engineering at Scale. They will discuss building trust and safety on consumer platforms, using operational AI to scale new markets, and the future of NFT marketplaces.

Join this session to learn more from Rajaraman about how NFTs will continue to foster transformation. He will discuss leading product in a Web3 startup environment, solving data problems with virtual identities and virtual worlds, and how AI is empowering creators with new capabilities. Before joining OpenSea, Karunamurthy served as Vice President of Commerce for Meta, Chief Technology Officer at WeWork, worked on products at Apple and Spotify, and worked in product management at Google and Twitter.

Graphics Processing Units (GPUs) are used for training artificial intelligence and deep learning models, particularly those related to ML inference use cases. However, using GPUs to deploy models at scale can create several challenges for ML practitioners. In this session, Varun Mohan, CEO and Co-Founder of Exafunction, will share the best practices he’s learned to build an architecture that optimizes GPUs for deep learning workloads. Mohan will explain the advantages for using GPUs for ML deployment, as well as where they might not have as many benefits. Mohan will also discuss cost, memory, and other factors in the GPU-vs-CPU equation. He will also discuss inefficiencies that may arise in different scenarios and some of the issues related to network bandwidth and egress. Mohan will offer techniques, including the importance of batching workloads and optimizing your models, to solve these problems. Finally, he will discuss how some companies are using GPUs to run their recommendation and serving systems. Before Exafunction, Mohan was a technical lead and senior manager at Nuro, where he saw the power of deep learning and the large challenges of productionizing it at scale.

It can be costly and time-consuming to figure out exactly the right data to train your model on. Scale Product Manager Bihan Jiang will describe an ideal Nucleus workflow to search for examples in your dataset, automatically tag similar examples, and sample the dataset for edge cases. She’ll then dig into label QA using models though charts and qualitative debugging. Lastly, she’ll show how to systematically evaluate models by discovering slices that encompass error cases for the model, as well as set up scenario tests to guard against model regressions. Plan to follow along in a public dataset like BDD100K (Berkeley DeepDrive) or apply her techniques to your own Scale dataset. Data curation prior to re-training your model and model testing and evaluation (T&E) should be a part of your modeling workflow. This workshop will focus primarily on computer vision applications and will be applicable to robotics, autonomous vehicle, logistics, manufacturing, and automated checkout use cases.

Aidan Gomez is the co-founder and CEO of Cohere, a provider of cutting-edge NLP models, and coauthor of "Attention is All You Need," one of the most-cited machine learning papers of all time, which introduced the world to the transformer architecture. Gomez and Scale CEO Alexandr Wang will discuss this paper's impact on the world and how it has led to the creation of large language models (LLMs) such as GPT-3 and BLOOM.

While the current generation of LLMs is impressive, Gomez will explain why they are not yet good enough for production applications and are inaccessible to most developers. He believes the industry must make it easy to access models and incorporate humans-in-the-loop to improve model performance so that the average developer without ML expertise can enrich their applications. In this fireside chat, Wang and Gomez will also discuss trends to watch for, including providing AI with tools such as knowledge bases to augment its capabilities and the advent of multimodal models with text, images, video, and audio fused in one model. Gomez will draw from his experience focusing on large-scale machine learning during his time at Google Brain, where he collaborated with many AI luminaries, including Geoff Hinton and Jeff Dean.

Despite international efforts to reduce deforestation, the world loses an area of forest that is equivalent to the size of 40 football fields every minute. Deforestation in the Amazon rainforest accounts for the largest share, contributing to reduced biodiversity, habitat loss for many of the world’s most threatened animals and insects, and more rapid climate change.

Satellite remote sensing offers a powerful tool to track changes in the Amazon. The Multimodal Learning for Earth and Environment Challenge (MultiEarth 2022) is the first competition aimed at monitoring and analyzing deforestation in the Amazon rainforest at any time and in any weather or lighting condition. Weather conditions in the Amazon are often humid and cloudy, making it difficult to gather clear images.

In this workshop, learn from Miriam Cha, Research Scientist in the Artificial Intelligence Technology Group at MIT Lincoln Laboratory, about multimodal representation learning for the earth and environment.

Cha will discuss the benefits and challenges of synthetic aperture radar (SAR), a sensor that transmits microwave signals and then receives back the signals that are returned from the earth’s surface. Although SAR images are extremely clear, they can be difficult for humans to interpret due to noise and lack of spatial correlations. This is why MultiEarth 2022 is using SAR with optical sensors for maximum effect.