Welcome to WednesdAI – Pixel Dreams’ weekly update with top stories from the rapidly evolving world of Artificial Intelligence.
This week’s news is potentially as revolutionary as the introduction of social media or the smartphone so let’s get right into it.
OpenAI Unveils Sora Video Generation
This week we’re covering possibly one of the biggest tech news stories of all time.
Last Thursday, OpenAI gave us our first look at Sora – their new AI video platform, boasting results that will change the way you think about what you see from here on out. The entire online world’s collective jaw dropped when we saw what OpenAI has been working on. No word yet on when we might have access to it. They’re taking the time to ensure safeguards are in place to prevent the use of Sora by bad actors. But AI-generated video has just taken a leap so huge that soon we will have to ask questions about everything we see on a screen.
Check out these examples from the official Sora announcement site:
Get the full details and see all the demo videos on OpenAI’s Sora website.
ChatGPT Will Remember You
In other OpenAI news, ChatGPT is getting a long-term memory function.
Users will be able to have ChatGPT recall user details and preferences across multiple chats. The feature will work across free and paid versions of ChatGPT. OpenAI says that sensitive information will not be stored in Memory, but policymakers and privacy advocates express concerns about how personal information is managed and protected.
Protecting Elections from A.I. Interference
Tech giants pledge to combat “deceptive artificial intelligence election content.”
Companies including Adobe, Google, Meta, Microsoft, OpenAI, and TikTok have pledged to an industry accord in anticipation of more elections taking place in 2024 than any year in history. Participating companies agreed to eight specific commitments at this past weekend’s Munich Security Conference. A draft of the accord aims to develop tools to “identify, label, and debunk AI-manipulated images and audio of public figures.”
See all eight commitments and read the full story at the Munich Security Conference website.
The App-free AI Smartphone
T-Mobile’s parent company is set to introduce a smartphone that replaces apps with an AI-based digital assistant.
At the upcoming Mobile World Congress in Barcelona, Deutsche Telecom (parent company of T-Mobile) will introduce “an app-free AI smartphone” that manages a variety of tasks through voice and text commands, eliminating the need for multiple apps. The company singles out travel planning, shopping, and multimedia editing as use cases that will be showcased at the conference. Cloud-based AI technology enables a high-quality user experience, even on entry-level smartphones while a high-performance version of an AI smartphone processes AI tasks on the device for faster, more energy-efficient operations tailored to individual users.
Up to 10x Faster than ChatGPT
AI startup Groq is making waves with its hardware innovations in the AI industry.
Groq’s specialized Language Processing Units (LPUs) enable nearly instantaneous responses for running Large Language Models (LLMs), significantly outperforming traditional GPUs with speeds of 500 tokens per second. This breakthrough, demonstrated to be about 4-10x times faster than GPT 3.5 in side-by-side comparisons, could revolutionize AI user experiences and offer a cost-effective alternative to GPUs. Groq’s advancements underscore the potential for new efficiencies and capabilities in AI technology.
Watch the comparison by Skill Leap AI, including Groq’s trademark lawsuit against X.com over their Grok chatbot:
See the side-by-side comparison on X.com and try Groq for yourself!
Join us every Wednesday for WednesdAI – a Pixel Dreams production
