This article digs into the ChatGPT conversation history of a 47-year-old Canadian man who was told by the bot that he’d “discovered a novel mathematical formula” that could take down the internet and create a force field vest.
When he realized this was delusional, he checked by asking Google Gemini.
The new Instagram Maps request to enable location services (feeding Meta valuable ad targeting data from your Android or iPhone) has spawned incredulous reactions, along with claims it’s on by default, despite Adam Mosseri’s denials.
Still, he says, “We’ll get out a few design improvements as quickly as possible.”
A Microsoft workers group says reporting by The Guardian, Local Call, and +972 Magazine “revealed incriminating details about Microsoft’s indispensable role as the technological backbone of Israel’s mass surveillance of Palestinians all over Palestine” with the IDF’s Unit 8200, despite the company’s denials.
...Microsoft and Unit 8200 worked closely to build a Microsoft-powered mass surveillance weapon that “collects and stores recordings of millions of mobile phone calls made each day by Palestinians in Gaza and the West Bank.”
Despite police stats showing violent crime in Washington D.C. is down 26 percent from last year, after dropping by 35 percent compared to the year before, the president has continued to insist that crime in the city is “totally out of control,” in posts and comments to reporters.
The latest comments from Trump and Elon Musk pushing to federalize the city follow a report that DOGE affiliate Edward Coristine and another person were attacked in a carjacking attempt that occurred between 3 and 4AM on August 3rd.
This Wired article shows how an indirect prompt injection attack against a Gemini-powered AI assistant could cause the bot to curse in responses and take over smart home controls by turning on the heat unexpectedly or opening blinds in response to saying “thanks.”
In a report dubbed “Invitation is all you need” (Sound familiar?), their Google Calendar invite passed instructions to the AI bot that were triggered by asking for a summary. Google was informed of the vulnerabilities they found in February and said it has already introduced “multiple fixes.”
If you also use 1Password and have an Android device, you may have also been dealing with autofill problems lately. However, updating to version 8.11.4 or higher and following the instructions to make sure Chrome is set to use third-party autofill seems to have worked on my phone to make the app more reliable again.
According to 1Password, the need to update its app and an uneven update rollout from Google caused a gap as Chrome switched from the old compatibility mode approach to the new native autofill experience.
[1password.community]
As legal wrangling over generative AI and copyright continues, ElevenLabs has launched its latest AI audio product with Eleven Music. There are many AI music generators out there, but the company claims this one “is cleared for nearly all commercial uses, from film and television to podcasts and social media videos, and from advertisements to gaming.”
However, except for podcasts, most of those uses are banned for most listed service plans.
As chatbots take a bigger bite out of search and the internet heads toward Google Zero, John Herrman writes, “The era of SEO is over. Say hello to GEO [Generative-engine optimization].”
The experts he spoke to indicated that publishers should focus on quality content to rank highly in AI search, but he also spoke to an ex-Googler who said he’s found evidence that the paid version of ChatGPT is using Google Search -- and he’s not the first one to say that.
[nymag.com]
