In light of the recent controversy surrounding the inaccurate and objectionable imagery generated by Gemini, an instance of a generative AI chatbot developed by Google, the organization has opted to reduce the functionality of its image generation capabilities until a resolution can be identified and operations can be reinstated.
Gemini’s Image Generation Capabilities Not as Strong as They Used to Be
Google explains that Gemini is a distinct product entity, which means it is not inextricably linked to its Search, AI models, or other services.
It makes sense that Google manually tuned the model once the app was complete so that the generated images would be diverse and not favor particular groups of people or ethnicities; this was done to ensure that users from different parts of the world could utilize the app without encountering significant difficulties.
None of the countermeasures, however, could have prevented what transpired thereafter. Gemini was incapable of accommodating use cases in which users specifically sought the presence of a particular ethnicity in an image, as per WCCFTech.
This resulted in the model being prompted with extremely peculiar images, including "A black teacher in a classroom," which were both inaccurate and offensive. Certain instances involved an outright refusal to produce images. Because the model erroneously perceived them as delicate when, in fact, they were not.
Google Pauses People Image Generation in Gemini, Promises Thorough Fixes
What is Google's position regarding this entire situation? In any case, the organization has disabled the capability to generate images of individuals in Gemini.
The organization has discussed how it will enable the feature once the problem has been resolved and the entire model has undergone rigorous testing.
Google has also acknowledged that AI models are not flawless; notwithstanding their advancements since their inception, they continue to be susceptible to hallucinations, which may give rise to complications.
Regarding the timeline, Google has not provided one; therefore, we will have to patiently await the completion of the Gemini image generation. We sincerely hope that when it returns, it will be significantly more capable of generating images than previously.
Photo: The Keyword/Google


Apple Names Steve Smith M&A Chief in Leadership Shakeup
Moonshot AI Executives Seek Settlement Ahead of Hong Kong IPO
Apple Cuts iPhone 18 Pro Production Orders Amid Weak Demand
Epic Games Delays Disney Shooter, Shifts to AI-First Development
OpenAI Annualized Revenue Hits $50 Billion, Below Earlier Estimates
Robinhood Chain Eyes Priority Gas Auctions for Faster Trades
TSMC Q3 Revenue Beats Estimates as AI Chip Demand Surges
Zcash Plans Quantum-Resistant Security Upgrade for January
Apple Supplier Luxshare Sees Limited Impact From US Patent Probe
OpenAI Revenue Hits $50 Billion, Below Earlier $70 Billion Estimate
Elon Musk Accuses Indian Telecom Giants of Blocking Starlink Launch
Arthur Hayes Announces FLOP Testnet Launch Within Weeks
Yandex Data Center Hit by Drone Strike as Russia-Ukraine Cyber Infrastructure Attacks Escalate
Nvidia’s China AI Chip Market Share Falls to 34% as Huawei Gains Ground
Oracle Uses Trucked Natural Gas to Power AI Data Centers Amid Pipeline Delays
Samsung, SK Hynix Shares Fall as China's CXMT Advances DRAM Technology 



