The bad news is AI has biases. The good news is we can fix them. I asked a specific model to generate an image of senior Asian executives sitting around a round table. I won't name the model, as the idea is to focus on the issue rather than criticize any company. The model
SAFETY
-
AI Murder Report: Critical Slides 38-41 Analysis
By
–
Hello Internet I’d like to report a murder. (See particularly slides 38-41.)
-
Whisper’s Hallucinations: Understanding ASR Limitations
By
–
We dive into the well defined task of automatic speech recognition (ASR), and describe why OpenAI’s Whisper, which has been integrated into ChatGPT, makes stuff up, or “hallucinates” as it’s called in the industry (bad nomenclature).
-
One-Model-Everything Approach Reduces ASR System Reliability
By
–
Their (and Muskrat and others’) quest to build one-model-for-everything has resulted in less reliable systems than what we’ve had before, even in the well defined task of ASR. As in, historically “hallucinations” weren’t problems in ASR systems!
-
DODGE Plan Risks Replacing Federal Experts With AI Tools
By
–
Imagine what will happen if DODGE replaces federal workers with these tools to perform all the tasks that expert federal workers do. We write “There is no “one weird trick” that removes experts & creates miracle machines that can do everything that humans can do, but better."
-
Why AI Speech Recognition Got Worse Despite Hype
By
–
In Scientific American, @asmelashteka & I ask: "In the age of unbridled AI hype, with the likes of Elon Musk claiming to build a “maximally truth-seeking AI,” how did we come to have less reliable speech recognition systems than we did before?"
-
Deep Understanding of Reality Carries Intrinsic Dual Use Implications
By
–
> Deep understanding of reality is intrinsically dual use. is a wonderful sentence (and broader insight).
-
Superhuman Client Quirks Exploitation Issues
By
–
It is difficult to tell whether that is happening because, again, what appears to be exploitation of Superhuman client quirks.
-
MSA Email Provider Phishing Attack URL Spoofing Threat
By
–
The named vendor is an MSA (email provider) that many, many founders will have an account with and the flavor text is update-your-credit-card, which many marks will immediately action because trusted vendor and appearance of an HTTPS link. I think they're doing some URL spoofing.
-
Chinese Developer Converts Unitree G1 Into Firefighter Humanoid Robot
By
–
Ce que j'attendais depuis un moment :
— VISION IA (@vision_ia) 14 avril 2025
Traduction : Genius, un développeur chinois, a modifié le Unitree G1 pour en faire un robot humanoïde pompier.
Nous avons besoin de plus de ces robots qui protègent des vies humaines. https://t.co/cXRpCZ2NGlCe que j'attendais depuis un moment : Traduction : Genius, un développeur chinois, a modifié le Unitree G1 pour en faire un robot humanoïde pompier. Nous avons besoin de plus de ces robots qui protègent des vies humaines.