🤔 FRIDAY FACTS
By the end of 2025, 95% of retail customer interactions are expected to be AI-assisted. But what’s the first product ever sold using AI recommendations?
Stick around for the answer! 👇️
🗞️ YOUR DAILY ROLLUP
Top Stories of the Day
📈 NVIDIA Hits $3.28T on AI Boom
NVIDIA’s valuation soared past $3.28 trillion in 2024, fueled by skyrocketing demand for AI-focused chips. It now trails only Apple, which neared $4 trillion amid AI-driven upgrades. Tech giants like Microsoft, Alphabet, and Amazon also thrived, boosting major indices—S&P 500 up 23.3% and Nasdaq up 28.6%. Analysts foresee continued growth in 2025, driven by AI investments and a potentially relaxed regulatory environment.
🩺 AI Chatbots Struggle in Diagnosing Patients
AI chatbots shine in structured medical exams but struggle in real-world patient interactions, achieving only 26% diagnostic accuracy, per a Harvard study. Models like GPT-4 excelled with summaries but faltered in open-ended conversations, exposing gaps in reasoning and data collection. While AI holds promise for clinical support, experts emphasize its inability to match human physicians’ holistic judgment and nuanced understanding of patient care.
⛏️ AI Mining Firm KoBold Hits $2.96B Valuation
KoBold Metals, an AI-driven mining startup backed by Gates and Bezos, reached a $2.96 billion valuation after securing $537 million in Series C funding. The company uses AI to locate critical minerals like lithium and copper, essential for EV batteries and AI centers. With investors like Andreessen Horowitz, KoBold aims to challenge China’s supply chain dominance, highlighted by a major copper find in Zambia and rising geopolitical tensions.
🥚 Smart Fridge Automates Grocery Shopping
Samsung has announced a partnership with Instacart to enhance its smart refrigerators with AI-driven grocery management. Utilizing Samsung's Vision AI technology, these fridges can monitor their contents, detect when items are running low, and suggest replenishments through Instacart's product matching API. Users can then order groceries directly from the fridge's touchscreen. This feature, aimed at streamlining meal planning and shopping, will be available via a firmware update later this year for models equipped with AI Vision Inside.
📧 PHISHING 2.0
AI-Generated Phishing Scams Target Executives with Hyper-Personalized Precision
The Recap: Corporate executives face a surge in hyper-personalized phishing attacks, fueled by AI’s ability to scrape vast online data and mimic authentic communication styles. As AI tools grow more advanced, cybersecurity experts warn of a rising tide of sophisticated scams that evade traditional defenses and carry costly consequences.
AI-powered phishing attacks leverage data scraped from online profiles to create convincing, personalized emails targeting executives.
British insurer Beazley and eBay report a sharp rise in cyberattacks using AI to mimic tone and style convincingly.
Generative AI tools make advanced phishing campaigns easier to execute, lowering the barrier for cybercriminals.
Researchers highlight AI’s role in crafting business email compromise scams, which have cost over $50 billion globally since 2013.
More than 90% of cyberattacks begin with phishing emails, and the average cost of a data breach rose to $4.9 million in 2024.
AI-generated scams evade detection by producing thousands of uniquely worded emails.
Security experts warn AI is not only scanning systems for vulnerabilities but also exploiting weaknesses in human behavior.
Forward Future Takeaways:
As AI-driven phishing attacks grow more sophisticated, businesses must adapt their defenses beyond traditional spam filters, focusing on AI-enhanced cybersecurity solutions and advanced threat detection systems. The rise of hyper-personalized scams underscores the urgent need for continuous employee training and robust data privacy policies to reduce vulnerability in an increasingly AI-powered threat landscape. → Read the full article here.
👾 FORWARD FUTURE ORIGINAL
Econ 10 | The Economics of Open-Source AI
In Econ 05 and Econ 06 I covered what is unique about software as opposed to the physical world. I fleetingly mentioned the concept of the ‘creative-commons free market of open source AI’. Why is it creative commons? How is it a free market? What exactly is open source and how does it matter to AI? Let’s explore these questions.
Let’s start with the very idea of open source. Software is written in a programming language. The resulting code - the set of instructions that define how the software should behave (such as with a cash machine (ATM) under what conditions it should trigger the dispensing of cash) - forms part of the ‘intellectual property’ of the people who built or own it.
Within the realm of intellectual property law, computer code has been deemed as something that is to be copyrighted rather than patented, just as other creative work in literary, educational and other spheres. This means, as hinted above, the builders or owners of the software retain the right to copy, distribute and otherwise make use of such code as they possess under the copyright.
However, the question of software code as copyrighted intellectual property took a different form early on. There arose this notion, and one that was spearheaded by stalwarts such as Richard Stallman, that code should be free to share, modify and reuse. This led, for example, to the concept of copyleft, as opposed to copyright of course, which provided for such freedoms, still within the legal framework of copyright law within the common law system. → Continue reading here.
🔬 RESEARCH PAPERS
CypherBench Boosts LLM Retrieval from Modern Knowledge Graphs
A new benchmark, CypherBench, aims to improve how large language models (LLMs) retrieve data from modern encyclopedic knowledge graphs like Wikidata. Researchers found that RDF-based graphs struggle with efficiency due to large schemas, overlapping relations, and complex identifiers that exceed LLM context windows. To address this, CypherBench introduces property graph views and uses Cypher queries for faster, more precise access to 7.8 million entities across 11 domains. The framework includes an RDF-to-property graph converter, a Cypher query generation pipeline, and new evaluation metrics—paving the way for better integration of knowledge graphs into AI systems. → Read the full paper here.
🛰️ NEWS
Looking Forward
🌊 xAI's Grok 3 Misses 2024 Launch: The anticipated release of xAI's Grok 3 model has been delayed, reflecting a broader trend of AI companies missing deadlines.
🤔 FRIDAY FACTS
The first product sold using AI recommendations? A book!
Amazon pioneered the use of AI for personalized recommendations in the late 1990s. Leveraging collaborative filtering, the platform began suggesting items based on users’ past purchases and browsing history. The result? Increased sales and a revolution in online shopping. Today, this technology fuels everything from movie suggestions to grocery delivery, proving that a good recommendation can go a long way—whether it's a bestseller or your nephew’s first novel.
Reply