- Starter AI
- Posts
- Google’s Gemini 1.5 Pro, A fresh take on Reinforced Learning, Apple’s latest features
Google’s Gemini 1.5 Pro, A fresh take on Reinforced Learning, Apple’s latest features
Controlling things with our eyes is possible!
Sthefania
May 16, 2024
In partnership with
Hello, Starters!
As we’re still catching up with days filled with updates, we keep wondering: how will AI transform our world in the upcoming year? Maybe it’s too soon to imagine, but there are endless possibilities!
Here’s what you’ll find today:
All about Google’s Gemini 1.5 Pro
Columbia + DeepMind’s approach to RL
Apple harnesses AI for new accessibility features
AI frenzy catches NASA's attention
Stability AI is facing challenges
And more.
🔎 All about Google’s Gemini 1.5 Pro (3 min)
All of Google's announcements during their I/O conference are worth a look, but let's focus on what Gemini 1.5 Pro offers. This flagship model now boasts a groundbreaking number of 2 million tokens, translating into almost 1.4 million words, 2 hours of video, or 22 hours of audio. We are talking about the largest input ever seen in a commercial model.
This uncovers a world of possibilities for developers working on demanding projects. Although Gemini 1.5 Pro is not available yet, there's a waitlist for those who want to try its power, and a version with 1 million tokens is coming up next month.
🧠 Columbia + DeepMind’s approach to RL (1 min)
A team of researchers from Columbia University and Google DeepMind has created a deep reinforcement learning model with predictive abilities. According to their study, the model's ability to solve tasks and anticipate changes in its environment based on its own actions helps it learn faster and better.
It's worth noting that this approach draws similarities from the way the hippocampus behaves in animals, positioning predictive learning as a key method to improve the development of models, making them more biologically realistic
Apple continues to share small insights into its AI journey, this time by announcing a series of accessibility features that will be released in the upcoming months. The company claims that these enhancements are made possible by Apple's hardware and software, highlighting their AI and machine learning capabilities, alongside the powerful Apple silicon chips.
Among the features are Music Haptics for users with hearing problems, Vocal Shortcuts, Vehicle Motion Cues to prevent motion sickness, and one of the most distinguishable, Eye Tracking. This groundbreaking method enables those with physical disabilities to control an iPad or iPhone using only their eyes, utilising a combination of AI, the front-facing camera, and on-device machine learning.
Learn How AI Impacts Strategy with MIT
As AI technology continues to advance, businesses are facing new challenges and opportunities across the board. Stay ahead of the curve by understanding how AI can impact your business strategy.
In the MIT Artificial Intelligence: Implications for Business Strategy online short course you’ll gain:
Practical knowledge and a foundational understanding of AI's current state
The ability to identify and leverage AI opportunities for organizational growth
A focus on the managerial rather than technical aspects of AI to prepare you for strategic decision making
🚀NASA is not unfamiliar with AI; however, as the technology becomes more prominent across various industries, they have decided to catch up with the trend by appointing their first-ever Chief AI Officer. To ensure they stay up-to-date on responsible development, David Salvagnini will take on this role.
😥Ups and downs have followed the path of Stability AI. Although the company has continued to offer its innovations to the world and achieved a billion-dollar valuation, the challenges are taking a toll. That's why they're currently in talks to sell due to a reduction in cash reserves.
⚡️Quick links
What did you think of today's newsletter? |