Welcome back AI enthusiasts!
In todayâs AI Report:
đłTech Giants Use YouTube Videos Without Consent?!
đOpenAI Founding Member Creates AI-Native School
đMicrosoftâs New SpreadsheetLLM
đ 5 Trending Tools
đ°Venture Capital Updates
đŒWhoâs Hiring?
Read Time: 3 minutes
đRECENT NEWS
TECH GIANTS
đłTech Giants Use YouTube Videos Without Consent?!
Image Source: Michael Tran/AFP/Getty Images/CNBC Screenshot
Tech Giants like Apple, Anthropic, Nvidia, and Salesforce used thousands of YouTube videos to train AI models without the creatorâs consent.
Key Details:
An investigation by Proof News found that Tech Giants used subtitles from 173,536 YouTube videos across 48,000 channels to curate datasets for AI model training.
Tech Giants purchased scraped content from these YouTube megastars:
Marques Brownlee: 19.1 Million Subscribers, Seven Videos Taken
Jacksepticeye: 30.7 Million Subscribers, 377 Videos Taken
PewDiePie: 111 Million Subscribers, 337 Videos Taken
MrBeast: 303 Million Subscribers, Two Videos Taken
Tech Giants purchased the scraped content from EleutherAI, a company that claims to be a ânon-profit AI research labâ that helps developers and researchers train AI models.
EleutherAI curated and organized the scraped content into datasets called âPile.â
âPileâsâ datasets are accessible and open to anyone on the internet with enough space and computing power to access them.
Why Itâs Important:
Tech Giants technically avoid âfaultâ because theyâre not directly scraping the content to curate datasets. EleutherAI would be at âfaultâ for breaking YouTubeâs Terms of Service.
Appleâs OpenELM, an efficient open-source language model family designed to run locally on iPhones, MacBooks, and iMacs, used âPileâ datasets for training.
đšSearch what YouTube videos were used to train AI models here.
đ©ș PULSE CHECK
Should Tech Giants compensate YouTubeâs content creators for helping them train AI models?
ANDREJ KARPATHY
đOpenAI Founding Member Creates AI-Native School
Image Source: Canva AI Image Generator
Andrej Karpathy, former Senior Director of AI at Tesla and founding OpenAI member, created an AI-native school with course materials supported and scaled by an AI Teaching Assistant.
Key Details:
Karpathyâs new venture, Eureka Labs, is an AI-native school that aims to provide a âTeacher + AI Symbiosisâ environment where human expert written course materials are supported and scaled with an AI Teaching Assistant.
The first class will be the âWorldâs Obviously Best AI Course,â or LLM101n. The course will help students train their own AI model.
Karpathy wants Eureka Labs to be âa proper, self-sustaining business,â but he also doesnât âwant to gatekeep educational content.â
Why Itâs Important:
Karpathy believes the ideal experience for learning anything new is under the guidance of subject matter experts who are âdeeply passionate, great at teaching, infinitely patient, and fluent in all of the worldâs languages.â
An AI Teaching Assistant can handle repetitive tasks like grading, explaining solutions, and offering personalized problems, freeing human experts to focus on more complex aspects of education like critical thinking.
AI RESEARCH
đMicrosoftâs New SpreadsheetLLM
Microsoft researchers introduced SpreadsheetLLM and SheetCompressor, new AI frameworks for encoding spreadsheets for Large Language Models (LLMS).
In other words, SpreadsheetLLM and SheetCompressor are experimental LLMs tailored to process, analyze, and execute information on spreadsheets like Microsoft Excel and Google Sheets.
Spreadsheets are characterized by their extensive two-dimensional grids, flexible layouts, and varied formatting options, which pose significant challenges for LLMs.
SpreadsheetLLM translates data in spreadsheets into a format that AI models can understand, allowing LLMs to analyze the data and help users write formulas or generate reports.
SheetCompressor identifies the most essential unique data points to create a more compact spreadsheet version.
SheetCompressor enables SpreadsheetLLM to work with much larger spreadsheets while requiring less computational power, making it more efficient.
đ TRENDING TOOLS
đCrayon tracks competitors to help you win deals.
đžElementor hosts, builds, and grows your dream website.
âPollThePeople effortlessly conducts consumer research.
đExplodingTopics surfaces rapidly growing topics before they take off.
đSequensAI generates content at scale to boost your lead pipeline effectively.
đźBrowse our always Up-To-Date AI Tools Database.
đ°VENTURE CAPITAL UPDATES
đŒWHOâS HIRING?
GuidewireRX (Remote): Machine Learning (ML) Intern, Fall 2024
Netflix (Los Angeles, CA): Machine Learning (ML) Intern, Research, Fall 2024
Jane Street (New York, NY): Software Engineer Intern, Summer 2025
Akuna Capital (Chicago, IL): Software Engineer Intern, Python, Summer 2025
NVIDIA (Santa Clara, CA): Developer Technology Engineer, AI, New Grad
đ€PROMPT OF THE DAY
PRICING
đPricing Strategies
Imagine youâre launching [Product/Service] for [Business] in [Industry]. Outline three potential pricing frameworks you could adopt, considering market competitiveness, perceived value, and long-term sustainability.
How would you balance initial user acquisition with maximizing revenue?
Product/Service = [Insert Here]
Business = [Insert Here]
Industry = [Insert Here]đFINAL NOTE
If you found this useful, follow us on Twitter or provide honest feedback below. It helps us improve our content.
How was todayâs newsletter?
â€ïžTAIP Review of the Day
âI look forward to reading The AI Pulse every day. I find it informative, educational, and bite-sized. Kudos to the team!â
REFER & EARN
đYour Friends Learn, You Earn!
{{rp_personalized_text}}
Refer 5 friends to enter đ°Julyâs $200 Gift Card Giveaway.
Copy and paste this link to others: {{rp_refer_url}}
