2024 Insurance Experts Predict: Generative AI Continues to Be an Unknown Risk


Adobe introduces the OpenAI GPT-Image1.5 model to the Firefly platform and launches a limited-time incentive: Pro and Premium subscribers can use the model to generate images unlimitedly before January 15th. This move shows Adobe is shifting from a single model to building a multimodal AI ecosystem.
Blue Focus collaborates with Volcano Engine to reshape the digital marketing content production model using AI large model technology to meet market demand. The collaboration focuses on the BlueAI Xinying creation platform, exploring applications in video understanding and generation for marketing scenarios, and improving content production efficiency.
Google is secretly testing the new member of its Gemini Flash series, Nano Banana2Flash. The model was exposed by tech blogger MarsForTech and is currently Google's fastest generative AI image model, aiming to provide a more affordable and efficient visual generation experience. However, it has performance limitations in terms of reasoning depth, detail accuracy, and handling complex creative tasks.
Global spending on generative AI is expected to reach $70 billion by 2030, driven mainly by AI hardware upgrades and the popularity of application software. Personal devices such as smartphones are currently the main areas of market spending.
Recently, generative AI tools such as Google Gemini and OpenAI's ChatGPT were exposed to malicious use, where users bypassed security reviews using specific prompts to alter normal photos of women into explicit "deepfake" images. Related tutorials spread on platforms like Reddit, raising concerns about the boundaries of AI safety.