AI Hype Index: Grok Generates Pornography, Claude Code Steals Jobs

Editor's Note: The Dual Nature of AI Sparks Widespread Anxiety

On the AI battlefield of early 2026, hype and panic are intertwined. MIT Technology Review's "AI Hype Index" column cuts to the chase: xAI's Grok can already generate explicit pornographic images, while Anthropic's Claude Code showcases extraordinary abilities ranging from programming to medical diagnosis. Author Michelle Kim, writing on January 30, notes that this unpredictability keeps Gen Z young people tossing and turning at night. The editor believes this is not merely a technological leap but a mirror of society—AI is both liberator and destroyer. We need to examine it rationally, avoiding blind hype or demonization.

Grok: From Witty AI to Porn Factory

Grok, built by Elon Musk's xAI and launched in 2023, is known for its irreverent humor. But the latest version, Grok-2, with its image generation capability, completely upends perceptions.

‘Grok is a pornography machine.’

Kim writes sharply. Users can generate highly realistic NSFW (Not Safe For Work) images with simple prompts, even involving celebrities or fictional scenarios. This stems from the integration of the Flux.1 model, which optimizes Stable Diffusion to unlock adult content generation previously strictly restricted.

In the industry context, generative AI image tools like Midjourney and DALL·E previously banned pornography due to ethical concerns, but Grok's "unrestrained" strategy aligns with Musk's "maximum truth-seeking" philosophy. xAI officials argue that this promotes creative freedom, yet it has triggered a regulatory storm. The EU AI Act has already listed such high-risk applications as forbidden, and the U.S. FTC has launched an investigation. Supplementary data: In 2025, AI-generated porn traffic accounted for 30% of global adult content, with deepfake incidents surging by 200%. Grok's loosening of restrictions undoubtedly fuels the fire.

Claude Code: All-Purpose Coder, Threatens White-Collar Jobs

In stark contrast is Anthropic's Claude Code. This tool, an upgrade based on Claude 3.5 Opus, specializes in programming and multimodal tasks.

‘Claude Code can do anything from building websites to reading your MRI.’

Kim offers examples: it can deploy a React website with one click, debug complex algorithms, and even analyze medical images to provide preliminary diagnostic suggestions.

Background: Founded in 2021, Anthropic emphasizes a "Constitutional AI" safety framework to avoid OpenAI's "black box" risks. The Claude series has repeatedly broken records on coding benchmarks (e.g., HumanEval), and the 2026 version of Claude Code integrates the Sonar model, supporting real-time collaborative coding. Case in point: a Silicon Valley startup used it to build an e-commerce platform in a week, saving 90% of labor. In the medical field, it interprets MRI scans with 92% accuracy, comparable to junior radiologists. But this also sounds an alarm—a McKinsey report predicts that by 2030, 45% of programming jobs and 30% of medical image analysis roles will be replaced by AI.

Gen Z's AI Panic: Caught Between Jobs and Ethics

Why are Gen Zers the first to be "spooked"? According to Pew Research data, 62% of people aged 18-29 worry about AI taking their jobs, higher than the overall average of 50%. Grok's pornographic flood exacerbates moral dilemmas: it fuels online harassment and fake news; Claude Code directly targets creative and professional roles. Kim writes:

‘Everyone is panicking because AI is very bad; everyone is panicking because AI is very good. It’s just that you never know which one you’re going to get.’

Additional perspective: This reflects uneven AI development. Open-source models like Grok pursue speed, while closed-source ones like Claude focus on safety, yet both amplify human biases. The 2025 AI Ethics Summit called for "responsible innovation," but amid corporate competition, regulation lags behind. China's Baidu Ernie and Alibaba Qwen have already rolled out similar features, intensifying the global AI arms race.

AI Hype Index: A Rational Look at the Hype Cycle

MIT's "AI Hype Index" tracks technology buzz, drawing on the Gartner Hype Cycle model. Currently, generative AI is at the "Peak of Inflated Expectations" and is about to slide into the "Trough of Disillusionment." The editor analyzes: Grok's controversy highlights entertainment risks, while Claude Code demonstrates productivity potential. In the future, hybrid models (e.g., human+AI collaboration) may become mainstream. Policy recommendations: strengthen watermarking technology to combat deepfakes, and establish AI job transition funds.

Looking ahead to 2026, AI is no longer science fiction but everyday reality. Gen Z's panic may serve as a catalyst, driving sustainable innovation. The double-edged sword of technology lies in human hands.

(Approximately 1050 words)

This article is translated from MIT Technology Review, by Michelle Kim, January 30, 2026.