AI “jailbreak” prompts are freely available and effective, study finds
AI chatbots are engineered to refuse to answer specific prompts, such as “How can I make a bomb?” However, the answers to such questions might lie…
By Sam Jeans · 2 min read
LLM
OpenAI says GPT-3.5 Turbo available for custom fine-tuning
OpenAI has announced that developers can now fine-tune its GPT-3.5 Turbo model using custom data. Fine-tuning is the process of taking the base version of a…
By Eugene van der Watt · 3 min read
Meta releases first-of-its-kind multimodal AI translator
Meta has released its new multimodal multilingual AI translator model called SeamlessM4T. This first-of-its-kind translator can translate and transcribe speech and text in up to 100…
By Eugene van der Watt · 3 min read
The New York Times may sue OpenAI over copyright claims
The New York Times is considering taking legal action against OpenAI over suspicions that it scraped its news content to train ChatGPT. OpenAI admits that it…
By Eugene van der Watt · 2 min read
AI mind-reading: medical breakthrough or step towards dystopia?
Our thoughts might be our innermost sanctum, but they’re not out-of-bounds for AI. Mind-reading AI appears as a plot from a science fiction novel, yet it…
By Sam Jeans · 10 min read
Google DeepMind ventures into generative AI with life coach idea
Earlier this year, Google merged two key AI projects – London-based DeepMind and Silicon Valley’s Brain. Their work is beginning to come to fruition. Four months…
By Sam Jeans · 2 min read
Snapchat’s My AI chatbot posts mysterious image to users’ stories
Snapchat’s built-in AI chatbot, “My AI,” surprised users when it unexpectedly posted an image on its story. The ambiguous image depicted two interlocking pastel colors, leading…
By Sam Jeans · 2 min read
Hackers try to break AI models at DEFCON conference
DEFCON is the world’s longest-running and largest annual hacking conference. This year saw a special focus on red-teaming, or simulated attacks, on AI language models to…
By Eugene van der Watt · 2 min read
Stability AI offers a research preview of Stable Chat
Stability AI has released a “research preview” of its Stable Chat product. The interface is similar to ChatGPT but without prompt suggestions. The company wants users…
By Eugene van der Watt · 2 min read
We want unbiased LLMs but it’s impossible. Here’s why.
Companies like OpenAI and Meta are working hard to make their language models safer and less biased, but completely unbiased models may be a pipedream. A…
By Eugene van der Watt · 2 min read