Getting Smarter vs Doing Better

Author

Nick Lipetzky
July 17, 2024 • Estimated Reading Time: 6 minutes

Hello! This is The Daily Spark! Your executive guide to navigating the AI maze with clarity and confidence. Here, we map, index, and signpost every twist and turn, ensuring you stay ahead in the ever-evolving landscape of artificial intelligence. Let's light the way together.

Getting Smarter vs Doing Better

"Don't just say you have read books. Show that through them you have learned to think better, to be a more discriminating and reflective person. Books are the training weights of the mind. They are very helpful, but it would be a bad mistake to suppose that one has made progress simply by having internalized their contents." - Epictetus

Today’s news shows the progress AI will make in five steps, how someone is using AI to help humans excel, and the potential progress for Excel (finally!). This is a core topic in AI, making it useful, not just intelligent. It’s been said many times that we don’t need technology to create our art, we need it to do the work keeping us from our art. Today we’ll chart an “AI Learning” course.

NEWS TO NAVIGATE AI

AI to teach AI to make AI

Former Tesla AI Director and OpenAI Cofounder, Andrej Karpathy, has started Eureka Labs to offer a new kind of school that integrates AI technology into the learning experience. Their flagship course, LLM101n, combines expert-created course materials with AI teaching assistants to provide an ideal learning experience. By leveraging generative AI, Eureka Labs aims to democratize education and make it accessible to anyone, anywhere.

Five Levels of AI

Based on leaked reports, OpenAI has introduced a new five-level system to track its progress toward developing artificial intelligence (AI) that can outperform humans. This framework was unveiled internally at an all-hands meeting and is intended to provide clarity on the company's advancements in AI development. Here's a breakdown of the five levels.

Rufus, your AI greeter

Amazon has launched a new AI shopping assistant named Rufus, which is now available to all users in the US through Amazon’s mobile app. Rufus is designed to enhance the shopping experience by providing quick and intelligent assistance directly within the app. Is anyone else picturing Rufus as an old man with a white mustache who can help you choose the right kind of grass seed for that patch of lawn that never grows?

SpreadsheetLLM

Microsoft researchers have introduced a research project called SpreadsheetLLM, which aims to enhance AI's interaction with an understanding of Excel spreadsheets. This research project addresses the complexities and challenges of applying AI to structured, logical, and formulaic datasets.

AI APPLIED
Research & Learning

Prompt with Purpose

Find Research Methodologies

Copy and paste into ChatGPT. Edit the [text inside these brackets] to fit your specific case.

Act as an academic research expert. Your task is to suggest appropriate methodologies for researching [topic]. Provide a comprehensive list of both qualitative and quantitative research methods that are best suited for the subject matter. Justify each methodology's relevance and potential advantages, ensuring they align with the research objectives. Additionally, address any potential limitations or challenges of each method, and offer potential solutions or alternative approaches. Your suggestions should be rooted in academic literature, ensuring their validity and appropriateness for academic research.

Human Side of AI

Enhance R&D Workflows

ResearchRabbit is a comprehensive research tool that aims to revolutionize the way researchers conduct their work. By reimagining literature search, alerts, and more, ResearchRabbit offers a range of features and benefits that can greatly enhance research workflows.

Value Creation

Customer Engagement via Education

AI-powered tools can be used for evaluating larger and more expensive AI aspirations. Start with small projects to test feasibility and assess impact. Use AI tools for writing assistance and video creation to prioritize initiatives and allocate resources wisely. This approach builds confidence, demonstrates value, and allows for adjustments based on real-world results. When the value is obvious and more investment makes sense then level up to creating a bespoke AI solution using a phased roadmap approach.

AI ELEMENTS
Machine Learning

Machine learning is like teaching a curious kid to ride a bike, but the kid is actually a computer.

Picture this: you have an enthusiastic learner named CompuKid. At first, CompuKid doesn't know how to balance or pedal straight. So, you start by giving CompuKid loads of examples - videos of people riding bikes, tips on posture and pedaling, and info about different terrains.

As CompuKid practices and processes all this bike-riding data, it starts to spot patterns and make connections. "Oh, leaning left turns the bike left!" or "Pedaling faster on hills helps keep the speed up!" Sure, CompuKid takes a few spills but learns from each one.

The more CompuKid practices and the more data you provide, the better it gets. Soon, CompuKid can tackle new routes, adjust to different bikes, and even pull off some tricks!

That's basically how machine learning works. We feed computers tons of data and let them practice solving problems or making predictions. Over time, they get better at the task, recognizing patterns and applying that knowledge to new situations - just like CompuKid nailing the art of cycling.

source: made with midjourney