AI Deepfakes, Data Grabs, Boundaries Pushed
09 April 2024: Tech Giants Bend Rules, TSMC's $11.6B Boost, AI Accessibility Woes, Musk: AI > humans by 2024, Meta labels AI content, JPMorgan's 400+ AI use case
Good Morning,
Latest in AI,
1. 1-Minute Roundup
2. Headlines
3. AI Investment Deals
4. Other AI News
1-Minute Roundup
OpenAI, Meta, and Google bend rules to gather AI training data, raising concerns.
TSMC to receive $11.6B in US grants and loans for Arizona AI chip plants
AI accessibility tools for blind users plagued by errors, lawsuits rise
JENA raises £1.2M in pre-seed funding for solopreneur tech stack
EnhencerAI secures $1M for AI-driven e-commerce advertising solutions
Lexverify raises £900K for AI risk management and compliance solutions
Prem Labs secures $14M to advance open-source AI technologies
Meta to expand AI content labelling ahead of the 2024 US election
Social media influencers face deepfake scams, targeting lesser-known creators
JPMorgan CEO Dimon compares AI's potential to the steam engine, identifies 400+ use cases
Alibaba Cloud reduces prices for global customers to gain an edge in AI and the cloud market
Elon Musk predicts AI will surpass human intelligence by the end of 2024
Headlines
OpenAI, Meta, Google Skirt Rules to Gather Data for AI Development
Major tech giants have been bending the rules and deploying questionable practices to obtain vast data to train their AI models.
Why It Matters: As AI advances, tech giants rapidly exhaust reputable online data sources. This has led to a desperate hunt for new data sources, potentially infringing on copyrights and user privacy.
Key Details:
OpenAI transcribed over 1 million hours of YouTube videos, possibly violating rules.
Google broadened its terms of service to tap user data from Google Docs and Maps.
Meta discussed buying Simon & Schuster for data and gathering copyrighted material.
Companies argue their actions fall under "fair use" laws
Creators and authors accuse AI firms of theft and demand compensation
The Big Picture: The race to lead in AI has driven tech companies to push legal and ethical boundaries in their quest for training data. As high-quality online data dwindles, some turn to "synthetic data" generated by AI. The outcome of ongoing lawsuits and copyright debates will shape the future of AI development and creators' rights in the digital age.
TSMC to Receive $11.6B in US Grants and Loans for Arizona AI Chip Plants
TSMC, the world's top chipmaker, will receive $6.6B in grants and up to $5B in loans from the US government to build a third factory in Arizona, expanding domestic production of critical technology.
Why It Matters: The funding is part of President Biden's effort to boost the US semiconductor industry under the Chips and Science Act. TSMC's investment will support emerging technologies like AI and national security.
AI Accessibility Tools for Blind Users Plagued by Errors, Lawsuits Rise
Blind internet users are struggling with error-prone AI-driven software installed by companies to comply with accessibility regulations, leading to lawsuits over poor website accessibility for disabled people.
Why It Matters: As businesses rely on AI-powered tools to meet accessibility requirements, the technology's limitations expose them to legal liabilities and hinder equal digital access for disabled users.
Zoom In:
AI-generated image descriptions on websites often mislead blind users
Lawsuits against companies' weak accessibility compliance up 13% in 2023
360,000+ companies worldwide use AI-powered accessibility tools
AI overlays can slow down websites and interfere with assistive tech
EU warns companies cannot rely solely on AI for compliance
The Big Picture: While AI accessibility tools are promoted as a quick fix, advocates say they often overpromise and under-deliver. Progress is expected as AI improves, but companies must be transparent about its limitations. Manual intervention remains necessary to ensure full accessibility compliance and equal digital access for disabled users.
AI Deals
JENA raised £1.2M in pre-seed funding to empower solopreneurs with an all-in-one tech stack that integrates a website, AI booking system, payments, CRM, marketing, and e-commerce.
EnhencerAI secured $1M in funding. The company aims to revolutionise AI-driven advertising for e-commerce businesses globally, focusing on ad analytics solutions and empowering businesses of all sizes.
Lexverify, an AI risk management assistant provider, has raised £900K in seed funding. After successful pilots in the financial services and pharmaceutical sectors, it aims to expand its proactive compliance solutions.
Prem Labs, an applied research lab focused on advancing open-source AI technologies, secured $14M in Seed funding. It aims to offer resources to train state-of-the-art generative AI models with full data sovereignty and privacy.
Others
Meta to Expand AI Content Labeling as 2024 US Election Approaches
Meta will apply "Made with AI" labels to a range of AI-generated videos, audio, and images on its platforms starting May 2024, extending its previous policy that only covered manipulated videos.
The move follows criticism from Meta's Oversight Board, which called the company's existing policies "incoherent" after an altered video of President Biden remained on Facebook despite not violating Meta's rules.
Social Media Influencers Face Deepfake Scams as AI Technology Advances
AI-generated deepfakes of social media influencers are being used for disinformation campaigns, product promotions, and non-consensual pornography, with lesser-known content creators, particularly women, becoming targets.
Start-ups like Ceartas and Alecto are developing AI-based solutions to help influencers identify and remove deepfakes. At the same time, social media platforms ban deliberately deceptive, manipulated media, copyright breaches, scams, and non-consensual nudity.
JP Morgan CEO Dimon Compares AI's Potential to the Steam Engine
Dimon believes AI could be as transformational as major technological inventions like the printing press, steam engine, electricity, and the internet, with the potential to "augment virtually every job."
JPMorgan has identified over 400 use cases for AI across various areas and is exploring the deployment of generative AI.
Alibaba Cloud Reduces Prices for Global Customers to Gain Edge in AI
Alibaba Cloud has reduced prices by an average of 23% across five core categories for international customers using its data centres outside mainland China.
The move aligns with Alibaba's "AI first strategy" to make core computing resources more accessible and gain a competitive edge against rivals like Amazon, Google, and Microsoft in the global cloud market.
Elon Musk Predicts AI Will Surpass Human Intelligence by End of 2024
Elon Musk believes AI will be smarter than individual humans by the end of 2024 and exceed the capabilities of all humans within the next five years as long as infrastructure can meet the technology's demands.
Like TheAIML?
Don’t keep it to yourself. Colleagues and friends can sign up here.
For even more,
Our newsletter archive: Click here
Follow us on Linkedin,
Follow us on 𝕏/Twitter
How are we doing? We would like to hear what you think about this newsletter. Let us know.