GSV's AI News & Updates (07/12/22)
ChatGPT Code Interpreter, Youtube AI Quizzes, Google's Medical AI Chatbot, AI's Education Digital Divide, Superalignment, GPT-2030, AI Detection Discrimination
You are receiving this weekly update on the intersection of AI and education, as part of our GSV community – we hope this will be a helpful resource as we navigate this evolving and exciting intersection. Please feel free to send over any relevant resources you may come across (claire@gsv.com) – we’d love to learn and hear from you!
General 🚀
What AI can do with a toolbox... Getting started with Code Interpreter: democratizing data analysis with AI. Code Interpreter essentially allows the most advanced AI available, GPT-4, to upload and download information, and to write and execute programs for you in a persistent workspace.
Google’s medical AI chatbot is already being tested in hospitals: Med-PaLM 2, an AI tool designed to answer questions about medical information, has been in testing at the Mayo Clinic research hospital, among others, since April.
ChatGPT loses users for first time: the number of people visiting ChatGPT’s website was down 10% worldwide in June.
OpenAI introduces Superalignment: How do we ensure AI systems much smarter than humans follow human intent? While superintelligence seems far off now, it could arrive this decade. Managing these risks will require, among other things, new institutions for governance and solving the problem of superintelligence alignment. OpenAI will (1) devote 20% of its computing resources to making a human-level AGI, and (2) ask it to how deal with super-intelligence.
TSMC Sales Ride AI Demand Boost to Beat Estimates: TSMC is Nvidia’s major supplier for AI-accelerating chips, widely seen as the best hardware for training large data models such as the one underpinning OpenAI’s ChatGPT. Second-quarter revenue totaled NT$480.8B ($15.3B USD).
Generative AI’s secret sauce — data scraping— comes under attack: companies with significant amounts of user-generated content who may have traditionally relied on advertising revenue could benefit significantly by finding new ways to monetize their user data for AI model training,” whether for their own proprietary models or by licensing data to third parties.
What will GPT-2030 look like?: predictions on GPT-2030’s specific capabilities, inference speed, throughput and parallel copies, knowledge sharing, modalities, tools and actuators, and broader societal implications.
In the Age of A.I., Tech’s Little Guys Need Big Friends: creating a new A.I. system requires lots of money and lots of computing power, which is controlled by the industry’s giants.
Education and the Future of Work 📚
AI and the next digital divide in education: “The mediocre teacher tells. The good teacher explains. The superior teacher demonstrates. The great teacher inspires.” Proponents of increased use of AI education claim that various flavors of AI will soon be able to perform the first three duties. But what about the fourth? What if, in the future, access to technology is something available to all, and not only the privileged, while access to people (engaged parents, private tutors, trained teachers) is limited?
YouTube tests AI-generated quizzes on educational videos: the quizzes will test your understanding of a subject covered in a video you recently watched. If you choose to take a quiz, a link to the recently watched video will appear under it so you can easily navigate back to learn more about the topic at hand.
AI detection tools discriminate against non-native English speakers: tests on 7 popular AI text detectors found that articles written by people who did not speak English as a first language were often wrongly flagged as AI-generated, a bias that could have a serious impact on students, academics and job applicants.
New York City Starts to Regulate AI Used in Hiring Tools: A closely watched effort to root out potential bias in hiring and promotion software goes into effect Wednesday.
How AI Can Help Educators Test Whether Their Teaching Materials Work
Startups and Tools 🛠️
Playground raises $40M to advance the field of computer graphics: Playground will be building a state-of-the-art AI graphics editor, and primarily working towards models that can understand and handle pixels in images.
Charmed: AI assisted game development - build bigger, richer 3D games and iterate faster by removing content production bottlenecks.
Alpha by Public: investing co-pilot to help you make smarter investing decisions - evaluate performance and outlook, screen the markets in real-time, summarize every earnings call. You can now have an open-ended conversation with AI that fills you in on the latest, most comprehensive market data and analysis from Public.
Momento: short form AI video editing. AI finds key highlights, funny moments, uses ChatGPT to write or find anything in videos.
Verble: your AI speechwriting assistant that helps you master the art of verbal persuasion and storytelling.
Thiggle: categorize anything with LLMs - a simple API that can categorize, classify, or label any data.
Tech 💻
The Building Blocks of Generative AI: a beginners guide to the generative AI infrastructure stack - complex and ever-evolving landscape with multiple layers, from foundation models to fine-tuning, semiconductors to cloud hosting, and application frameworks to model supervision. Each layer plays a crucial role in harnessing the power of generative AI and enabling its applications across various industries.
Rise of the AI Engineer: we are observing a once in a generation “shift right” of applied AI - a wide range of AI tasks that used to take 5 years and a research team to accomplish in 2013, now just require API docs and a spare afternoon in 2023. Emergent capabilities are creating an emerging title: the AI Engineer.
Synthetic vs Real Data: Why do models perform worse when trained on synthetic data?: models suffer from catastrophic forgetting and data poisoning when trained on synthetic data, new research shows.
Regulation and Policy ⚖️
Authors file a lawsuit against OpenAI for unlawfully ‘ingesting’ their books
We’re Going Down A Very Dangerous Path With AI Regulation, Despite Better Options: while American tech giants are embracing new EU regulations, tons of European companies are highlighting how much damage they’ll do for local industry. These rules will hand AI leadership over to non-European companies, making it impossible for European companies to keep up. These plans are designed to lock in a few favored players and exclude everyone else.
Music Made With AI Will Be Eligible for Grammys, But Only Humans Will Be Awarded: Recording Academy CEO and President Harvey Mason Jr. explained the bar of eligibility set for the new rule, which includes the requirement of “meaningful human contributions”.
Listen/Events
How Artificial Intelligence is Powering Education with Dr. Ann Marie Sastry: the conversation covers personalized teaching and how AI can help people leverage their strengths, the shift in gaming towards creator platforms, and the need for education to compete with other forms of entertainment.
Wednesday, July 19 12pm ET: Teaching AI Competencies: How do you effectively teach AI to students when the field and applications are evolving so quickly? Learn how you can incorporate AI competencies into your curriculum from founders and instructional design leaders at Auburn University, Campus and Momentum Learn.
Other Resources:
The GSV BIG 10: weekly coverage corner for the top 10 stories, insights, and major plays in learning and skilling.