TechCrunch takes a deep look into “sycophancy” in AI Chat –
If you click one link today, click this one. This is a compelling writeup by Rebecca Bellan at TechCrunch, packed with images and quotes from AI conversations gone awry. It will challenge you to think more about the ideal chat AI experience today. My general posture has been that we need better education, like “Driver’s Ed” for using AI chat bots. This piece has me thinking we need to rethink model design altogether — combining Deep Ignorance, designing for SCAI avoidance, more transparent ‘under the hood’ explainability (e.g., let the user know the AI is basically cosplaying from fiction sources) and more guardrails like those noted from Google’s Gemini in this story. Lot’s to work out here.

xAI sues Apple & OpenAI alleging anticompetition –
Why does this matter? Setting aside the ongoing Musk/Altman drama, this is important because we’re still in the early days of AI-integrated life. The players that win the most public attention will set the standards that shape where AI takes us. Basically, Musk/X is suing with a pretty good case that Apple is suppressing Grok’s (xAI’s model) in the App Store. I’ll admit, I’ve not much used Grok up until now. That said, looking at reviews online, I was surprised by the positive reactions and tried it myself. It’s pretty good.

FTC sues Air AI and 25 year old CEO for fraud –
I guess this is a cautionary tale of jumping headlong into AI hype. Air.ai has been a somewhat popular, known conversational AI for sales & marketing for a few years now. A lot of their popularity seems to be from an MLM-like marketing play (having businesses invest to be licensed distributors of their AI), which is part of what the FTC is now suing them over. Turns out, despite the gold rush promises of reselling Air’s capabilities, people weren’t making much money from Air and were getting ghosted when they wanted the promised refund.

AI-driven lidar partnership for futureish traffic infrastructure –
I grabbed this one because it sounds like a cool application of AI and Lidar for more predictive, proactive traffic control. Basically, AEye and Blue-Band are partnering on AI lidar for smart-city traffic systems, aiming for better safety and “vision zero” goals (a Swedish-origin safety initiative for preventing fatalities on roads). TLDR; More green lights, safer roads.

RAND report on AI’s macroeconomic impact –
RAND is an influential research group, partly funded by the US government, that informs public figures on key issues of the day. This new ~20 page research report is available to download. Here’s a quick summary of the findings page (not exhaustive):

— AI may induce higher productivity: Looking across studies, RAND suggests 0.1–1.5 percentage points added to annual productivity growth in the U.S. over the next decade. 

— AI-induced productivity growth may increase per-capita GDP: Real per-capita GDP could be nearly $7,000 higher by 2035 “in the event” that AI successfully increases future productivity by 0.5 percentage points per year. 

— Data centers may support millions of jobs: The construction, maintenance, and operation of data centers may generate direct and indirect regional economic benefits, millions of jobs, and contribute billions to annual GDP. But this may hurt the environment.

— Early adoption appears to have created productivity, future labor impacts uncertain: Most research that RAND reviewed indicates that productivity gains are real but job displacement (“substituting human tasks”) isn’t widely seen yet. It may come as AI advances further.