#019 Midweek Special - GPU-Rich vs GPU-Poor, Meet Yahoo’s New AI Assistant, The Dark Side of AI Development.
Fresh & Hot curated AI happenings in one snack. Never miss a byte 🍔
This snack byte will take approx 7 minutes to consume.
AI BYTE # 1 📢 : GPU-Rich v/s GPU-Poor: The Emerging Gap In The Tech World
⭐ The tech industry is witnessing a new kind of divide: the GPU-rich and the GPU-poor.
GPUs, or graphics processing units, are specialized chips that can perform complex calculations faster than CPUs, or central processing units.
GPUs are essential for powering applications such as artificial intelligence, gaming, virtual reality, and cryptocurrency mining.
However, not all tech companies have equal access to GPUs.
Some companies, such as Nvidia, AMD, Google, Microsoft, and Amazon, have invested heavily in developing and acquiring GPUs, giving them a competitive edge in the market.
These companies are the GPU-rich, and they can offer superior products and services to their customers and partners.
On the other hand, some companies, such as Intel, IBM, Oracle, and SAP, have lagged behind in the GPU race. These companies are the GPU-poor, and they face challenges in keeping up with the demand and innovation of the GPU-rich.
The GPU-poor may have to rely on renting GPUs from cloud providers or partnering with GPU-rich companies to access their technology.
The GPU divide has implications for the future of the tech industry. The GPU-rich may be able to dominate the fields of AI, gaming, VR, and crypto, while the GPU-poor may struggle to catch up or diversify their offerings.
The GPU divide may also affect the talent pool, as more developers and engineers may prefer to work for the GPU-rich companies.
The GPU divide is not a permanent state of affairs. The GPU-poor may still have opportunities to close the gap by investing in their own GPU development or acquiring GPU startups.
The GPU-rich may also face competition from new entrants or regulators who may challenge their dominance.
The GPU divide is a dynamic and evolving phenomenon that reflects the changing landscape of the tech world.
AI BYTE # 2 📢 : Meet Yahoo’s New AI Assistant: A Smart and Convenient Way to Manage Your Email
⭐ Yahoo has unveiled a suite of AI tools that can enhance your email experience and help you with tasks like finding information, saving money, and writing messages.
The tools are available as an opt-in beta for Yahoo Mail users, including those who use Yahoo’s platform to access email from other providers like Google or Microsoft.
One of the new features is an AI assistant that can answer questions based on your inbox and automatically draft suggested emails to merchants when you forget to use your gift cards or discount codes.
Instead of searching for a sender name or string of text to find an email that tells you how much you owe on a bill or when your next flight departs, the AI-enhanced search box lets you ask explicit questions with the help of an autocomplete box.
The assistant is powered by large language models (LLMs), which are advanced AI systems that can generate realistic text and images. Yahoo’s AI can also pop up quick summaries of existing emails, highlighting key information and suggesting actions like adding events to a calendar.
Another new feature, called Shopping Saver, is designed to detect cases where users buy something online without realizing they have an applicable gift card, store credit, or discount code.
In that case, Yahoo’s AI will detect the new purchase and unused savings opportunity and pop up an alert to suggest sending an email to the merchant asking to retroactively use the gift card or discount. If the user agrees, a new AI writing assistant will automatically draft the first version of the email.
The writing assistant can be used for other purposes as well, drafting emails based on a quick prompt from the user and an AI scan of previous correspondence with the recipient to get a sense of context and appropriate tone.
Users are free to suggest the AI adjust the tone and they can edit the draft themselves before clicking the send button.
Yahoo says the tools do entail its AI parsing your email, but it takes pains to ensure the content of your messages stays private and emphasizes that email contents aren’t used to train AI models accessible by other users.
The tools will likely continue to evolve as AI advances and continues to get better at understanding user intentions.
If you are interested in trying out these new features, you can opt-in by clicking on the “Try Beta” button in your Yahoo Mail account.
You can also opt-out at any time if you change your mind. Yahoo hopes that these tools will give users of other email providers a reason to give Yahoo another look and make their email experience more convenient and enjoyable.
AI BYTE # 3 📢 : The Dark Side of AI Development: Low Wages and Poor Conditions for Overseas Workers
⭐ Artificial intelligence (AI) is a booming industry that promises to revolutionize various fields and sectors.
However, behind the scenes, there is a dirty secret that many AI companies are not talking about: they rely on millions of overseas workers to train their algorithms and make them work.
These workers, mostly from the Philippines, are tasked with labeling images, text, and other data that feed into AI systems. This process is essential for AI to learn and understand the world, but it is also tedious, repetitive, and low-skilled.
The workers are often employed by platforms such as Remotasks, which is owned by San Francisco-based startup Scale AI.
Scale AI provides data annotation services to major tech companies such as Airbnb, Pinterest, and Uber. Scale AI claims to have over 10,000 workers on Remotasks, but according to data and interviews obtained by the Washington Post, the company has often failed to pay them on time or at all.
Poor Wages and Working Conditions
The workers on Remotasks are freelancers who sign up online and take tests to qualify for different tasks. They are paid by the task, which can range from a few cents to a few dollars. However, the tasks are often complex and time-consuming, requiring high accuracy and attention to detail.
According to the Washington Post, many workers have complained that they were not paid what they were promised, or that they were not paid at all. Some workers said they spent hours or days on a project, only to receive a fraction of the expected amount. Others said they were not paid for months, or that their accounts were deactivated without explanation.
The workers have no effective way to appeal or complain, as they have no contracts or legal protection. They also have no access to benefits or social security.
Some workers said they had to work long hours, sometimes overnight, to make ends meet. Some said they suffered from eye strain, headaches, and stress.
Ethical and Social Implications
The situation of overseas workers raises ethical and social questions about the AI industry. Some experts have called these workplaces “digital sweatshops”, where workers are exploited and dehumanized.
They argue that AI companies should be more transparent and accountable for their data sources and practices and that they should ensure fair wages and working conditions for their workers.
The situation also reflects the global inequality and power imbalance in the AI field. While AI companies in the US and other developed countries reap huge profits and benefits from their products and services, they outsource labor-intensive and low-value tasks to workers in developing countries, who receive little compensation and recognition.
This creates a “race to the bottom”, where workers compete for low-paying jobs with no security or dignity.
The situation also poses challenges for public awareness and education about AI. Many people may not realize how much human labor and intervention is involved in creating and maintaining AI systems.
They may also not be aware of the potential risks and harms that AI can cause, such as bias, discrimination, privacy violations, and social disruption.
There is a need for more public engagement and dialogue about the ethical and social implications of AI development.
Choose Wisely