30 秒でわかる
Matthew Berman looks at fictional AI-generated images of public figures and tests image prompts in Grok 2. He characterizes the tool as uncensored and contrasts it with DALL·E, but the few examples do not establish either system's complete safety rules or current behavior. The images are synthetic, not evidence of real events.
- 1A 2024 image toolBerman introduces Grok 2's image generation and calls it uncensored. The brief demonstration is not a comprehensive safety test.
- 2Fictional public-figure imagesHe reviews generated images depicting public figures in invented situations; they are not real photographs of the events portrayed.
- 3Two example promptsBerman tries prompts involving Elon Musk and Mark Zuckerberg and contrasts the results with his understanding of DALL·E.
ポイント
- Berman examines fictional generated images of public figures and submits two further prompts.
- He describes Grok's 2024 behavior as less restrictive than DALL·E in these examples.
- A small set of prompts does not establish that an image system has no safeguards.
- Generated images of real people should not be treated as records of actual events.
チャプターごとに整えた全文を読む
Generated examples
Berman introduces Grok 2 and reviews fictional images of public figures.
Matthew Berman introduces Grok 2's image generation and looks at fictional images involving public figures, including Taylor Swift, Donald Trump and Kamala Harris. These are synthetic depictions, not evidence that the portrayed events occurred.
Trying prompts
He submits two prompts and compares the experience with DALL·E. The sample is too small to establish complete policy differences.
Berman then requests an invented image of Elon Musk and another based on Mark Zuckerberg's public style, and comments on the results. He contrasts the experience with DALL·E in ChatGPT. His description of Grok as completely uncensored is a judgment based on a small set of 2024 examples, not proof that no safeguards exist or a description of current policy.
同じチャンネルの記事
34:21EN
Matthew Berman· 2026年10月27日Matthew Berman Examines Anthropic’s IPO Filing, AI Growth and GovernanceMatthew Berman, Sam Altman · Forward FutureThe presenter reviews the S-1’s financial metrics, compute strategy and voting structure while questioning the tension between rapid AI development and safety.
0:26ENMatthew Berman· 2026年10月25日Sam Altman on Human Values and Problem-Solving in an AI WorldSam Altman · OpenAIIn a short interview excerpt, Altman considers the enjoyment of solving hard problems and the human role in interpreting results.
0:36ENMatthew Berman· 2026年10月25日Kimi K3.1 and a Rubik’s Cube Solver DemonstrationMatthew Berman · Forward FutureMatthew Berman shows a scrambled virtual cube being solved and gives the Kimi K3.1 test a favorable assessment.
0:21ENMatthew Berman· 2026年10月24日An AI-Generated Kimi Benchmark Chart Used Invented ScoresMatthew Berman · Forward FutureMatthew Berman asks for a graph of Kimi K3.1 Pro benchmarks, then discovers that the displayed figures are illustrative placeholders.
25:50ENMatthew Berman· 2026年10月24日Kimi K3.1 Pro Tested: Strong 3D Demos, Uneven Results and Competitive PricingMatthew Berman · Forward FutureMatthew Berman tests the open-weight model on coding, 3D rendering and visual explanations, then compares benchmark results and prices.
32:20ENMatthew Berman· 2026年10月23日Who Is Local AI For? The Costs, Trade-offs and a Hybrid FutureMatthew Berman · Forward FutureMatthew Berman weighs data control and independence against hardware cost, setup complexity and inference speed.