- AI Breakfast
- Posts
- What percentage of Americans think AI is sentient?
What percentage of Americans think AI is sentient?
Good morning. It’s Friday, July 19th.
Did you know: On this day in 1968, Intel was founded.
In today’s email:
GPT-4o Mini
Deepmind’s Open-Source AI
20% of Humans think AI is already sentient
5 New AI Tools
Latest AI Research Papers
You read. We listen. Let us know what you think by replying to this email.
Today’s trending AI news stories
OpenAI just dropped GPT-4o mini — here’s what we know about this cheaper and faster AI
OpenAI has released GPT-4o mini, a scaled-down version of its GPT-4o model, designed to replace GPT-3.5 Turbo. This new model enhances core GPT-4o capabilities, offering improved speed and affordability while integrating safety features such as Instruction Hierarchy to guard against malicious prompts. Initially focused on text, GPT-4o mini will also support image processing, speech-to-speech, and voice guidance.
The model is now available to all ChatGPT users, with enterprise access expected next week. GPT-4o mini delivers a 82% score on multitask language understanding, compared to 70% for GPT-3.5 Turbo, and is over 60% cheaper. It features a 128K context window, advanced multilingual support, and default safety measures. The model supports high-throughput scenarios and flexible, global pay-as-you-go deployment. Azure AI complements this launch with performance upgrades and extensive data residency options across 27 regions, enhancing its global accessibility and efficiency. Read more.
Related story: OpenAI reportedly holding talks with Broadcom and others to develop new AI server chip
Google Deepmind develops open-source AI to tackle biases in evaluating language models
Researchers from Google DeepMind, Google, and UMass Amherst have introduced FLAMe (Foundational Large Autorater Models), a new open-source AI system designed to enhance the evaluation of AI-generated text.
FLAMe, equipped with 12 billion parameters and trained on over 5.3 million human ratings from 102 diverse tasks, outperforms commercial models such as GPT-4 and Claude 3 in 8 out of 12 evaluation tasks, particularly in factual accuracy and attribution. This model is noted for its reduced bias, thanks to its training on publicly available, open-license data. A variant, FLAMe-RM, optimized for reward modeling, achieved 87.8 percent accuracy in the RewardBench test, surpassing other models.
The researchers highlight FLAMe’s role in promoting transparency and openness in AI text evaluation while acknowledging risks such as potential neglect of human perspectives and amplification of existing biases. FLAMe and its training data will be made publicly available to support further research. Read more.
Many people think AI is already sentient - and that's a big problem
A recent survey reveals widespread misconceptions about AI among the US public. Approximately 20% of respondents believe AI systems are already sentient, while 30% think artificial general intelligences (AGIs) capable of performing any human task are in existence. Conducted by Jacy Reese Anthis and colleagues, the survey highlights that perceptions of AI as sentient are influenced by hype and anthropomorphism in media and marketing.
Anthis suggests that such misconceptions could lead to misplaced trust in AI systems, especially as their use expands into sensitive areas like government and policing.
The survey underscores the need for clearer communication about AI’s true nature and limitations to prevent potential overreliance and misinterpretation of AI’s capabilities. (Paywall) Read more.
Etcetera: Stories you may have missed
5 new AI-powered tools from around the web
Scade.pro effortlessly integrates audio, image, text, and video AI models with its no-code platform, enhancing projects by up to 90% in speed and cost.
Paillor is an employee support and automation platform that unifies knowledge bases, SaaS automation, and help desk functionalities with AI integration.
Flow Studio generates 3-minute, film-quality videos with consistent characters, engaging plots, matched background music, and sound effects, eliminating post-production editing.
Miro’s Intelligent Canvas integrates AI and custom workflows with tools, featuring AI Sidekicks, Intelligent templates, and multiplayer tools like polling and timelines. Spaces consolidates all project content into a central hub.
Archie uses AI to design and plan software applications in minutes, generating requirements, designs, architecture, plans, and SOWs for enhanced productivity.
arXiv is a free online library where researchers share pre-publication papers.
Thank you for reading today’s edition.
Your feedback is valuable. Respond to this email and tell us how you think we could add more value to this newsletter.
Interested in reaching smart readers like you? To become an AI Breakfast sponsor, apply here.