Skip to main content

Asking Chatbots For Short Answers Can Increase Hallucinations, Study Finds

2 months ago
Requesting concise answers from AI chatbots significantly increases their tendency to hallucinate, according to new research from Paris-based AI testing company Giskard. The study found that leading models -- including OpenAI's GPT-4o, Mistral Large, and Anthropic's Claude 3.7 Sonnet -- sacrifice factual accuracy when instructed to keep responses short. "When forced to keep it short, models consistently choose brevity over accuracy," Giskard researchers noted, explaining that models lack sufficient "space" to acknowledge false premises and offer proper rebuttals. Even seemingly innocuous prompts like "be concise" can undermine a model's ability to debunk misinformation.

Read more of this story at Slashdot.

msmash

Google Launches New Initiative To Back Startups Building AI

2 months ago
Google has launched the AI Futures Fund, a new initiative to invest in AI startups that are building with the latest tools from Google DeepMind. TechCrunch reports: The fund will back startups from seed to late stage and will offer varying degrees of support, including allowing founders to have early access to Google AI models from DeepMind, the ability to work with Google experts from DeepMind and Google Labs, and Google Cloud credits. Some startups will also have the opportunity to receive direct investment from Google. "The AI Futures Fund doesn't follow a batch or cohort model," a Google spokesperson told TechCrunch. "Instead, we consider opportunities on a rolling basis -- there's no fixed application window or deadline. When we come across companies that align with the fund's thesis, we may choose to invest. We're not announcing a specific fund size at this time, and check sizes vary based on the company's stage and needs -- typically early to mid-stage, with flexibility for later-stage opportunities as well." Startups can apply here.

Read more of this story at Slashdot.

BeauHD