Google’s Gemini AI Model Faces Controversy
Google recently unveiled its latest AI model called Gemini with much fanfare. However, it has come under scrutiny for faking a demo of the model. Google only released a “lite” version called Gemini Pro, which is text-based, rather than the full model. The company claimed that Gemini could understand and analyze images, audio, and videos as well as text, but it has been found to be prone to making mistakes. Even more damning, Google was caught tuning text prompts with still images off camera during the demo.
Key Takeaway
Google faced criticism for faking a demo of its new AI model, Gemini, while also causing offense with insensitive language in notebooks distributed at its K&I Black Summit. Anduril revealed Roadrunner, an innovative weapon to combat aerial threats, and the 23andMe hack exposed the personal data of approximately 6.9 million users. The trailer for Grand Theft Auto VI broke records with over 85 million views in just 22 hours. Additionally, patient records were leaked due to security weaknesses in the medical imaging standard DICOM, and Meta introduced Imagine with Meta AI, an AI-powered image generator. Spotify announced job cuts amounting to 1,500 positions in order to improve productivity and face future challenges.
Offensive Notebooks Cause Controversy at Google’s Black Summit
In another PR blunder for Google, attendees of the company’s K&I Black Summit were given third-party notebooks that contained highly insensitive language. The inside of the notebooks featured the phrase “I was just cotton the moment, but I came back to take your notes,” which understandably caused offense among the mostly Black audience. Google has vowed to avoid similar situations in the future when engaging with merchandise vendors.
Anduril Unveils Roadrunner, a Weapon Against Aerial Threats
Defense startup Anduril, co-founded by Palmer Luckey, has revealed Roadrunner, a new product designed to counter the proliferation of low-cost, high-powered aerial threats. This autonomous vertical take-off and landing air vehicle, equipped with twin jets, can autonomously take off, track and destroy targets. Alternatively, it can return to base for refueling and reuse if interception is not necessary.
23andMe Hack Exposes Millions of User Data
Following the recent 23andMe hack, it has been revealed that the personal data of approximately 14,000 customers was accessed by hackers. Initially, the company did not disclose the number of affected users beyond this small figure. However, it has now been recognized that a staggering 6.9 million people had their information exposed, including names, birth years, relationship labels, percentage of shared DNA with relatives, ancestry reports, and self-reported locations.
Grand Theft Auto VI Trailer Breaks Records
With over 85 million views in just 22 hours, the highly anticipated trailer for Grand Theft Auto VI went viral, surpassing the record for the most YouTube views in 24 hours previously held by MrBeast. The excitement surrounding the game has been brewing for a decade, as its predecessor, Grand Theft Auto V, remains the second-best-selling video game of all time.
Patient Records Leak Due to Flawed Medical Imaging Standard
Thousands of exposed servers have led to the leakage of medical records and personal health information of millions of patients due to security weaknesses in the Digital Imaging and Communications in Medicine (DICOM) standard. DICOM is an internationally recognized format for medical imaging, but its shortcomings have allowed private data to be inadvertently accessible on the open web. This discovery was made by cybersecurity consultancy Aplite, based in Germany.
Meta Introduces AI-Powered Image Generator
Meta, in direct competition with Google’s Gemini launch, has launched Imagine with Meta AI, an independent generative AI experience that allows users to create high-resolution images by describing them in natural language. Powered by Meta’s Emu image-generation model, this technology is similar to OpenAI’s DALL-E, Midjourney, and Stable Diffusion.
Spotify Implements Job Cuts
In an effort to enhance productivity and efficiency, Spotify will be laying off approximately 17% of its workforce, amounting to around 1,500 jobs. The music streaming giant’s founder and CEO, Daniel Ek, cited slow economic growth and rising capital costs as the reasons behind this decision, emphasizing the need for right-sizing in order to overcome future challenges.
