IMD: ““LLMs are programmed to project authority while preying on the insecurities of users,” says [Amar] Bhidé. “Their responses are at once utterly confident and aim to affirm users’ opinions or desires.” If a CEO asked an LLM a series of questions about business strategy, the replies would more than likely align with the CEO’s existing views, assuring them that their thinking is sound and that they are on the right course. “They play courtier, not devil’s advocate,” continues Bhidé. “LLMs seem to have been designed to flatter the user, thereby encouraging their continued use. ’Oh, what a brilliant idea!’ they will effectively say, convincing the user that they are really smart.”” [via Arnold Kling]
BCG Newsletter: “The way companies typically treat a capital investment is very different from how they treat a human investment. Agentic AI has characteristics of both. Like capital, agentic AI systems require substantial investment upfront in development, integration, and infrastructure. Like a human workforce, they also require ongoing variable costs for evaluation, tuning, domain adaptation, usage-driven inference, and governance. This blend of capex and opex makes agentic AI an economic hybrid, fundamentally reshaping enterprise investment models.”
Fei-Fei Li: “Marble is a frontier model. What’s really remarkable is it generates a 3D world at a simple prompt. That prompt could be, Give me a modern-looking kitchen. Or, Here’s a picture of a modern-looking kitchen. Make it a 3D world. The ability to create a 3D world is fundamental to humans. I hope one day it’s fundamental to AI. If you are a designer or architect, you can use this 3D world to ideate, to design. If you are a game developer, you can use it to obtain these 3D worlds, so that you can design games. If you want to do robotic simulation, these worlds will become very useful as training data for robots or evaluation data. If you want to create immersive, educational experiences in AR [or] VR, this model would help you to do that.”
Bloomberg: “Since the Covid-19 pandemic, a new generation of traders has flooded markets through apps that blend brokerage, betting and social media antics. They’re using tools built for speed, stakes and engagement: stock options with zero days to expire (0DTE) that can deliver thousand-percent swings in minutes; leveraged exchange-traded funds, or ETFs, that triple the pain or pleasure of a daily move; event contracts that let users wager on the consumer price index, earnings calls or NFL games; memecoins and tokenized stocks. More than half of the S&P 500’s daily options volume now comes from 0DTEs, instruments that barely existed on any scale five years ago. Assets in leveraged ETFs have soared sixfold since the onset of the pandemic, to $240 billion. Sports event contracts, essentially a form of gambling, clocked $507 million in trades on Kalshi, one of the biggest prediction markets, during just the NFL’s opening week this season. Day in and day out, Wall Street, which likes to talk about managing risk, has been manufacturing new ways to take it. More assets to trade. More chances to win. More dopamine.”
