Revolutionizing Women's Productivity with Project Lunaris + News

Anna Barrow
December 05, 2023

TODAY’S AGENDA

IBM and Meta’s AI Alliance: Unpacking the Collaboration and Its Impact on Open-Source AI
The Energy Footprint of AI Images: A Surprising Comparison to Smartphone Charging
Empowering Women with Project Lunaris: A Revolutionary App for Aligning Productivity with Lunar Cycles
Jargon Decryption: Neural Networks
AI-Generated Images: Showcasing the power and creativity of Midjourney AI in generating stunning visual content from v5-showcases.

Figure 1: AI Alliance | Source: GPT-4 DALL-E Generated Image

Formation of the AI Alliance:
- Collaboration between IBM, Meta, and over 45 organizations.
- Aim: To promote open-source AI and open innovation.
Goals and Initiatives:
- Develop trust and validation metrics for AI.
- Focus on AI training, hardware, and open-source models.
- Establish project standards and guidelines.
Diverse Membership:
- Involves tech giants like AMD, Intel; research labs like CERN.
- Participation from leading universities and AI startups.
- Covers sectors like healthcare, silicon, and software-as-a-service.
Collaboration and Sharing:
- Emphasizes faster, more inclusive innovation.
- Aims to identify and mitigate risks before product release.
Impact on the AI Industry:
- Encourages a more open, collaborative approach to AI development.
- Stands in contrast to closed, proprietary AI models.
- Democratizes AI development, spreading benefits across industries.

Figure 2: AI Image Footprint | Source: GPT-4 DALL-E Generated Image

Energy Consumption of AI Images:
- Generating one AI image = Energy used to fully charge a smartphone.
Comparison of AI Tasks:
- Image generation is the most energy-intensive AI task.
- Text generation is less energy-intensive.
- Example: Generating 1,000 images = Carbon emissions of driving 4.1 miles.
Impact of Multi-Purpose Models:
- Large, versatile AI models consume more energy.
- Specialized models for specific tasks are more energy-efficient.
Cumulative Environmental Impact:
- Small energy usage adds up with millions of daily AI users.
- Calls for conscious use of AI to reduce environmental footprint.

Figure 3: Project Lunaris | Source: GPT-4 DALL-E Generated Image

What is Project Lunaris?
- An innovative app designed for women.
- Helps track the 28-day lunar cycle.
- Aims to align productivity with hormonal cycles.
Why Project Lunaris?
- Challenges the 24-hour solar cycle norm.
- Recognizes women's unique physiological rhythms.
- Empower women by working with their natural cycles.
Key Features:
- User-friendly mobile and web app dashboards.
- Personalized cycle tracking and productivity insights.
- Tailored strategies to enhance daily effectiveness.
Empowerment Through Technology:
- Shifts from male-centric productivity models.
- Promotes understanding of female biological rhythms.
- Offers a more inclusive approach to productivity.
Vision and Goals:
- Empower women to harness their natural cycles.
- Foster a balanced and fulfilling lifestyle.
- Drive a conversation about inclusive tech solutions.
Broader Impact:
- Breaks the mold of one-size-fits-all solutions.
- Encourages more empathetic and personalized tech development.
- Sets a precedent for gender-sensitive technology.
Looking Ahead:
- Project Lunaris is more than an app – it's a movement.
- A step towards tech that understands and respects diversity.
- A tool for women to achieve enhanced productivity and well-being.

Figure 4: Neural Networks | Source: GPT-4 DALL-E Generated Image

What is a Neural Network?
- Think of it as a computer program designed to learn and make decisions like a human brain.
- It's a simulation made with ordinary computer parts, not an actual brain.
- Neural networks can recognize patterns and make decisions without being explicitly programmed.
How Does a Neural Network Work?
- It's made of layers of artificial neurons, called units.
- Units receive information (input units), process it, and then give a response (output units).
- Between input and output are hidden units, that form the network's core.
- Units are connected by 'weights' that determine how much one unit influences another, like brain cells connected by synapses.
Simple vs. Deep Neural Networks
- Simple neural networks have a few layers, good for basic tasks.
- Deep Neural Networks (DNNs) have many layers, that handle complex problems.
- DNNs need more examples to learn from, making them suited for complex tasks.
- Despite their complexity, they are still just algorithms, not real brains.

Imagine a bunch of tiny lights (units) inside a box (the computer).
Each light can either be on or off, depending on the lights it's connected to.
Some lights get information from outside (input units), and some show what the box is thinking (output units).
The lights in the middle (hidden units) help figure out what the outside information means.
The strength of the connections (weights) determines which lights affect others more.
A DNN is like having many boxes inside each other, for really tough problems.