#ai-reliability

#ai-reliability

When Baker McKenzie announced plans to cut roughly 700 business services staff, the firm pegged some of the blame squarely on AI, feeding into the technology-fueled paranoia surrounding artificial intelligence. It certainly doesn't help that the AI hype cycle turns on tech gurus juicing up their VC sugar daddies with promises that AI will replace human workers any moment now,

Grok's track record is spotty at best. But even by the very low standards of xAI, its failure in the aftermath of the tragic mass shooting at Bondi Beach in Australia is shocking. The AI chatbot has repeatedly misidentified 43-year-old Ahmed al Ahmed, the man who heroically disarmed one of the shooters, and claimed the verified video of his deed was something else entirely - including that it was an old viral video of a man climbing a tree.

A judge is racing to break up Google's advertising empire before they can appeal, while Microsoft's Copilot stumbles on camera. Australia's sweeping social bans, Roblox's selfie requirement, and flawed AI moderation spark sharp debate on what happens when online gatekeeping gets serious. Top MAGA Influencers Accidentally Unmasked as Foreign Trolls NetChoice Sues Virginia To Block Its One-Hour Social Media Limit For Kids Roblox is requiring 9yo kids to submit a video selfie to prove age

Intuit said its tools, such as TurboTax, Credit Karma, QuickBooks and Mailchimp, will be accessible through ChatGPT, allowing users to ask questions and complete tasks such as estimating tax refunds, reviewing credit options, or managing business finances. With users' permission, Intuit's apps will be able to access their financial data to generate responses and complete tasks, like sending marketing messages or issuing invoice reminders.

Which? surveyed more than 4,000 UK adults about their use of AI and also put 40 questions around consumer issues such as health, finance, and travel to six bots - ChatGPT, Google Gemini, Gemini AI Overview, Copilot, Meta AI, and Perplexity. Things did not go well. Meta's AI answered correctly just over 50 percent of the time in the tests, while the most widely used AI tool, ChatGPT, came second from bottom at 64 percent. Perplexity came top at 71 percent. While different questions might yield different results, the conclusion is clear: AI tools don't always come up with the correct answer.

I use it for legal advice. So when I am needing to know the answer to a question, I'll take a picture and snap it and, like, put it in there,

AI hype is following a well-worn path. During the dot-com boom, we were promised the internet would bring an overnight revolution. While it was revolutionary, some changes arrived quickly, but most unfolded over years, marked as much by failures and false starts as by lasting breakthroughs. I was at Inbound in San Francisco last week, and the AI hype was overwhelming. Conversations were either focused solely on the tactical use of AI or, interestingly, on reframing AI in our minds.