GSV's AI News & Updates (10/15/24)
Meta MovieGen, Zoom AI Avatars, Amazon AI Shopping Agents, Nevada AI Controversy, Career Karma OutRival, AI Job App Bots, Synthetic Data, Adobe Content Authenticity
Join me at the leading edge of all things AI in EDU at the second annual AI SHOW @ ASU+GSV, where 12,000+ PreK to Gray attendees will experience cutting-edge AI innovations through live EdTech demos, promptathons and hackathons, hands-on workshops, incredible thought leadership, and so much more — April 5-7 at the San Diego Convention Center. Register for FREE now, capacity is limited.
General 🚀
OpenAI Projections Imply Losses Tripling to $14B in 2026: OpenAI allegedly projects reaching profitability in 2029 at $100B of revenue, with $44B of cumulative losses to get there. Microsoft’s investment terms give it 20% of OpenAI’s revenue.
OpenAI launches new ‘Canvas’ ChatGPT interface tailored to writing and coding projects: Several consumer AI providers are converging around editable workspaces as a practical way to use generative AI. ChatGPT’s new interface offers similar features to Anthropic’s Artifacts, launched in June, and the viral coding companion, Cursor.
Meta announces Movie Gen, an AI-powered video generator: The models (30B video parameters and 13B audio parameters) can generate up to 16-second videos complete with sound effects, background music, and even personalized characters. The outputs look similar to OpenAI's Sora, and it’s unclear if or when Meta plans to release the model publicly.
Google’s Grip on Search Slips as TikTok and AI Startup Mount Challenge: A flurry of new offerings for advertisers is hitting the market. AI is reshaping how search works, with Google, Microsoft, and Perplexity introducing ads in AI-generated responses. Google’s U.S. search ad market share is expected to drop below 50% for the first time in over a decade.
Zoom will let AI avatars talk to your team for you: You’ll soon be able to create a custom AI avatar of yourself that you can use to record and send short messages. Zoom’s CEO envisions AI avatars as “digital twins” that can attend meetings or handle tasks like emails, allowing users to focus on other activities.
Amazon Dreams of AI Agents That Do the Shopping for You: Amazon has introduced AI-generated shopping guides and is exploring autonomous AI shopping agents, but concerns remain about granting these agents too much control, like making irreversible purchases. Amazon's AI is fine-tuned to be a better "shopping concierge".
Anthropic co-founder Dario Amodei on How AI Could Transform the World for the Better: Dario discusses transformative changes and how they can be practically implemented across: biology and physical health, neuroscience and mental health, economic development and poverty alleviation, peace and governance, work and purpose.
AI startup Writer, currently fundraising at a $1.9B valuation, launches new model to compete with OpenAI: San Francisco-based Writer launched a large AI model using cost-effective synthetic data, spending $700,000 on training compared to competitors' millions. Now raising up to $200 million at a $1.9 billion valuation, experts warn this approach may affect model performance and amplify biases.
Education and the Future of Work 📚
Nevada Asked A.I. Which Students Need Help. The Answer Caused an Outcry.: Nevada's recent implementation of an AI system to identify at-risk students has sparked significant controversy and raised complex ethical questions about the use of artificial intelligence in education. The AI-driven approach, developed by an external contractor, dramatically reduced the number of students classified as "at-risk" in Nevada from over 200,000 to under 65,000.
Parents Sue School That Gave Bad Grade to Student Who Used AI to Complete Assignment: The lawsuit alleges the district unfairly targeted the student by denying him National Honor Society admission while allowing another AI-using student to join, underscoring growing tensions around academic integrity, discipline, and fairness in AI use in education.
This District Is Building an AI Chatbot—But Not for the Students: Developing the chatbot in-house allows the district to customize responses to staff needs, like providing advice on curriculum and values, and ensures better privacy since there’s no third-party involvement in handling data.
Career Karma founders launch OutRival to help companies build AI agents: OutRival aims to enhance customer service by helping companies build AI agents that work alongside human teams, not replace them. It focuses on scaling personalized experiences through AI, especially in areas like customer support. OutRival is leveraging leftover capital from the $40M Series B round Career Karma raised in 2022.
‘I Applied to 2,843 Roles’: The Rise of AI-Powered Job Application Bots: A trending GitHub project, Auto_Jobs_Applier_AIHawk, uses AI to automate job applications on LinkedIn, with one user applying to 150 jobs overnight. While critics argue that it reduces human interaction and accelerates hiring automation, the tool appeals to job seekers frustrated with lengthy recruitment processes.
Accenture To Train 30,000 Staff On Nvidia AI Tech In Blockbuster Deal: This training effort is tied to the formation of a new Accenture Nvidia Business Group, which aims to accelerate enterprise adoption of AI, particularly generative AI and "agentic AI systems"
Gloat Revolutionizes Workforce Agility with New AI-Powered Products: has They have launched two innovative products as part of its Agile Workforce Operating System: Skills Planner and Studio. Skills Planner provides HR leaders and workforce planners with real-time insights into skill gaps and transferability across the organization. Studio empowers HR and L&D leaders to quickly build and launch customized talent programs.
The University of Texas at Austin releases new AI teaching and learning tools: an AI-powered writing assistant, personalized learning pathways, and a virtual teaching assistant, to enhance teaching and learning experiences on campus
Chan Zuckerberg Initiative Commits Funding To Help Educators Shape How AI Will Be Used in Classrooms: The grants will support organizations like Education First for AI adoption, ISTE for piloting a national initiative reaching 90,000 educators, and Leanlab Education for expanding co-design infrastructure for K-12 educators.
The world needs a ‘premortem’ on generative AI and its use in education: Rather than wait for its widespread adoption to study the effects of this technology, the Center for Universal Education has launched a task force aimed at anticipating its effects on education.
Sticky humans in a post-AGI world: I simply don't believe that any near-future AI, no matter the prompt, could long-term take the social place of real intellectual interaction with an adult human or a community of humans. The details of what was taught is usually secondary, after all, often forgotten or inconsequential; it is the aesthetics and modes of intellectual engagement that remains.
Tech 💻
The promise and perils of synthetic data: As real-world data becomes scarcer and more expensive, companies like Anthropic, Meta, and OpenAI are increasingly using AI-generated synthetic data to train their models. Synthetic data is vulnerable to the same biases and errors as the data used to generate it, leading to potential over-reliance on poor-quality or biased data.
Microsoft may have an audio-to-image generator in the works, new patent shows: The system would capture live audio, convert it into a text transcript, summarize it using a large language model (LLM), and then generate real-time images based on the spoken content. This process would run continuously, providing a dynamic visual representation of the conversation or presentation as it unfolds. The audio-to-image generator could revolutionize fields like education, business, entertainment, and accessibility by generating real-time visual aids for lectures, meetings, podcasts, live performances, and for individuals who benefit from additional sensory input.
Anthropic challenges OpenAI with affordable batch processing: The new Message Batches API offers significant cost savings and efficiency improvements for businesses processing large volumes of data with AI models (allows submission of up to 10,000 queries in a single batch).
OpenAI Unveils Secret Meta Prompt—And It’s Very Different From Anthropic's Approach: Under the hood, OpenAI’s structured approach to prompt generation sets it apart from Anthropic’s human-like, chatbot approach.
Meta’s Yann LeCun says worries about AI’s existential threat are ‘complete B.S.’: LeCun argued that today’s large language models lack some key cat-level capabilities, like persistent memory, reasoning, planning, and an understanding of the physical world.
TikTok parent company ByteDance has a tool that's scraping the web 25 times faster than OpenAI
Safety and Regulation ⚖️
Adobe has a new tool to protect artists’ work from AI: The free Content Authenticity web app allows creators to apply attribution and ‘do not train’ tags to any image, video, or audio work.
In South Korea, deepfakes wreck women’s lives and deepen gender conflict: Police detained 387 people for deepfake crimes this year, with the majority being teenagers. The true scale is believed to be much larger, with South Korean women making up more than half of the people featured.
Wikipedia Declares War on AI Slop: A collaborative effort launched by Wikipedia editors to combat the increasing problem of unsourced, poorly-written AI-generated content on the platform.
Silicon Valley is debating if AI weapons should be allowed to decide to kill
Other
A Shift in the World of Science: What this year’s Nobels can teach us about science and humanity. The article raises concerns about whether future scientific breakthroughs will be driven by AI rather than human understanding, as AI becomes the tool crafting revolutionary work.
Tesla CEO Elon Musk unveils robotaxi and robovan in self-driving car splash: