- AIML Newsletter
- Posts
- AI Shockers & Game-Changers: Must-See Revelations! - Feb 8, 2024
AI Shockers & Game-Changers: Must-See Revelations! - Feb 8, 2024
From Billion-Dollar Bets to AI That Reads Your Mind – Dive In!

Buckle up for a whirlwind tour of the AI universe in today's edition of the AIML newsletter. We're serving up a feast of revelations that'll make your jaw drop, from Vanguard's eye-watering investment in AI-driven funds to KLM's cheeky AI that predicts your no-show at dinner. LinkedIn's got an AI genie out of the bottle, promising to turbocharge your job hunt, while Microsoft's Copilot is the new creative sidekick you didn't know you needed. And that's just for starters. Get ready to be dazzled by the smarts of AI as we explore its latest escapades across the globe.
Business / Finance
AI's Quiet Conquest: Vanguard's Billion-Dollar Bet
Vanguard has integrated machine learning into $13 billion of its quant stock funds, moving beyond its traditional investment strategies.
The funds have seen positive early results, with some beating benchmarks and peers indicating the potential of AI in improving investment outcomes.
This shift towards AI includes using neural nets and factor strategies for nuanced stock predictions; it marks a significant move for the traditionally conservative investment firm. Read
Bright Skies: KLM's AI Meal Guessing Game Cuts Waste
KLM uses AI to predict which passengers will miss their flights, reducing meal loads and cutting food waste by 63% in a trial.
The AI, developed by Kickstart AI, adjusts meal numbers based on predictions from 17 days to 20 minutes before departure, saving about 2.5 meals per long-haul flight.
This initiative is part of KLM's broader use of AI for operational efficiency and sustainability. Read
Job Hunt, Upgraded: LinkedIn's AI Assistant Steps In
LinkedIn introduces AI-powered tools for premium users, using OpenAI's technology to suggest job fits and profile enhancements.
The AI, supported by OpenAI and indicated by a sparkle emoji under job listings on Linkedin, offers advice through a chat interface, analysing user profiles against job requirements.
Designed to ease the job search process, the AI can also interpret lengthy job descriptions, offering insights into company culture, benefits, and more. Read
Designing the Future: Microsoft's Copilot Gets a Creative Boost
Microsoft Copilot introduces enhanced AI features, including a refined Deucalion model, offering a more streamlined user experience with a fresh look and suggestive prompts.
The platform now includes advanced editing capabilities for images, allowing users to modify colours, blur backgrounds, or change styles, with additional features available for Copilot Pro subscribers.
Copilot is set to expand with Designer GPT, providing a dedicated space for users to bring their creative ideas to life visually. Read
Regulatory

Safeguarding the Future: OpenAI's Child-Friendly Initiative [In Response to EU’s CSA Regulation]
OpenAI has established a Child Safety team to explore ways to make AI tools safe and appropriate for children, focusing on preventing misuse or abuse by underage users.
This initiative aligns with efforts to comply with child protection laws and includes hiring a child safety enforcement specialist to manage sensitive content and enforce OpenAI's policies.
OpenAI is also collaborating with Common Sense Media to develop AI guidelines suitable for children, reflecting a proactive approach to addressing the potential risks and ethical concerns of AI usage by minors. Read
OpenAI’s GPT Trademark Turbulence: USPTO's Final Verdict
The US Patent and Trademark Office (USPTO) has issued a final refusal to OpenAI's application for a GPT trademark, labelling it as a generic term that merely describes the technology's features.
This decision stems from the view that GPT, short for Generative Pre-trained Transformer, directly relates to the functions of OpenAI's AI and ML services, making it descriptive rather than distinctive.
OpenAI faces the choice of appealing this decision to the Trademark Trial and Appeal Board or finding alternative means to protect its use of the GPT term. Read
Health
Heartbeats and Bytes: AI's New Rhythm in Healthcare
Researchers developed an AI framework to classify heart diseases from audio signals, showcasing the potential for non-invasive diagnostics.
Machine and deep learning techniques such as Convolutional Neural Networks (CNNs) and Support Vector Machines (SVMs), among others, were employed to analyse heart sound signals, achieving high accuracy in detecting heart conditions.
The study highlights the effectiveness of audio signal-based diagnosis, presenting a promising avenue for enhancing medical diagnostics and patient care. Read
Brain Teaser 🧠

What is my Lucky Number?
I am thinking of a three-digit number. The second digit is four times as big as the third digit, while the first digit is three less than the second digit. What is the number?
AIMO Prize: The $10 Million AI Math Challenge
XTX Markets has established a $10 million fund for the Artificial Intelligence Mathematical Olympiad Prize to advance AI's capability in mathematical problem-solving.
The grand prize of $5mn will be awarded to the first publicly shared AI model to enter an AIMO-approved competition and perform at a standard equivalent to a gold medal in the IMO.
The initiative invites global participation, with the inaugural competitions set to begin in early 2024, leading up to a presentation at the 65th International Mathematical Olympiad in Bath, England. Read
Morning Chuckle 😄
Why did the AI break up with its algorithm?
Because it found out it was seeing other data!
Did You Know 🤔
Guinea Pigs for AI: Did you know that the game of Chess was one of the first platforms to test Artificial Intelligence? In 1997, IBM's AI system "Deep Blue" made history by defeating the reigning world champion, Garry Kasparov, in a six-game match. This monumental event marked the first time a reigning world champion lost a game to a computer under standard chess tournament conditions and served as a significant milestone in the development of AI.
Quote of The Day ✨
"Artificial Intelligence is the new electricity." - Andrew Ng
Just as electricity transformed industries one hundred years ago, AI is now set to revolutionise our world in ways we're just beginning to imagine.
Security
UK’s Co-op's Counter-Crime Crusade: AI and Armor
The Co-op is bolstering its supermarkets with AI surveillance and 200 secure tills in response to a 44% spike in retail crime, equating to 1,000 daily incidents.
The UK retailer's new security investments, totalling £200m, also include locked cabinets for high-value items and a specialised undercover security team, despite facing nearly 298,000 shoplifting incidents last year.
These measures aim to curb organised theft and assaults on staff, with the Co-op advocating for stricter legal repercussions for retail worker attacks. Read
Chromebooks at a Crossroads: Denmark's Privacy Stand
Denmark's Datatilsysnet demands more privacy guarantees for using Google services in schools, citing improper use of student data by Google.
Municipalities have until March 1 to align with the new privacy directive, potentially leading to a Chromebook phase-out by August 1.
The concern includes Google's use of data for performance analytics and AI feature development, which is a challenge under European privacy laws. Read
Research
Smaug-72B: The Open-Source AI Monarch
Smaug-72B, a new open-source language model by Abacus AI, tops the Hugging Face rankings, outperforming notable models like GPT-3.5 and Mistral Medium.
Originating as an enhanced version of Alibaba's Qwen-72B, Smaug-72B shines in multiple natural language processing benchmarks.
It's the first open-source model to achieve an average score above 80 across critical evaluations, marking a significant milestone in AI accessibility and performance. Read
Coding Made Easy: Introducing Taipy for Data Wizards
Taipy is a Python framework designed to help data scientists and ML engineers develop full-stack applications without learning additional programming languages.
It simplifies application development by integrating data and AI algorithms into user-friendly web applications with pre-built components and robust data management features.
Taipy is ideal for complex applications like demand forecasting, offering tools for version management and pipeline orchestration to facilitate collaborative work. Read
OLMo Unveiled: Shining a Light on NLP's Black Box
OLMo (Open Language Model) is a new AI framework by the Allen Institute for AI, designed to enhance transparency in NLP (Natural Language Processing).
It offers the model's weights and inference capabilities and the entire toolkit used for its development, including training codes, datasets, and architecture documentation.
OLMo aims to foster an open research ecosystem, providing comprehensive evaluation tools and supporting various applications, from simple language tasks to complex generative ones. Read
Unscrolling History: The Digital Decipher

The Vesuvius Challenge 2023 awarded a $700,000 Grand Prize for successfully reading a 2000-year-old scroll using AI and machine learning, revealing ancient texts on music, food, and enjoyment of life.
A global effort involving computer vision and machine learning cracked the code, allowing the ancient texts buried by the eruption of Mount Vesuvius to be read without physically unrolling the scrolls.
The winning team's innovative approach combined multiple model architectures, including a TimeSformer-based model, to achieve unambiguous and readable text from the carbonised scrolls. Also, Github code. Read
And there you have it – a rollercoaster ride through the AI wonderland, where machine learning meets billion-dollar gambles, and algorithms predict everything from your flight flakiness to your next career leap. We've seen AI become a sustainability hero with KLM, a job hunter's best buddy on LinkedIn, and push creative boundaries with Microsoft's Copilot. Not to mention the brainy moves in healthcare, the ethical leaps for child safety, and the open-source triumphs reshaping the AI landscape. If today's tales of AI magic and wonder have tickled your fancy, why keep the fun to yourself? Spread the word and let your friends in on the secrets of the AI universe!