Top Must-Read GenAI Articles and Authors
Spotlighting the best 10 GenAI articles, as their knowledge have been a gift to us this past year. 🎁
Hello Engineering Leaders and AI Enthusiasts!
We wish you a very Happy Holidays! ☃️🎄🎅
The holiday season is upon us, and in the spirit of giving, we wanted to spotlight 10 of the best GenAI minds to follow.
Enjoy our hand-picked holiday reading list - it's sure to delight any GenAI enthusiast.
In today’s edition:
🎥 Tutorial: How to make and share custom GPTs by Charlie Guo🚀 Startup productivity in the age of AI by jason@calacanis.com
💡 Practical Tips for Finetuning LLMs Using LoRA by Sebastian Raschka, PhD
🔧 The Interface Era of AI by Nathan Lambert🧮 “Math is hard” — if you are an LLM – and why that matters by Gary Marcus
🎯 OpenAI's alignment problem by Casey Newton
👔 In Praise of Boring AI by Ethan Mollick🎭 How to create consistent characters in Midjourney by Linus Ekenstam
📱 The Mobile Revolution vs. The AI Revolution by Rex Woodbury
Let’s go!
Why Incumbents LOVE AI
Since the release of ChatGPT, we have seen an explosion of startups like Jasper, Writer AI, Stability AI, and more.
Far from it: Adobe released Firefly, Intercom launched Fin, heck even Coca-Cola embraced stable diffusion and made a freaking incredible ad (below)!
So why are incumbents and enterprises able to move so quickly? Here are some brief thoughts on it by Shomik Ghosh
LLMs are not a new platform: Unlike massive tech AND org shifts like Mobile or Cloud, adopting AI doesn't entail a massive tech or organizational overhaul. It is an enablement shift (with data enterprises already have).
Talent retention is hard…except when AI is involved: AI is a retention tool. For incumbents, the best thing to happen is to be able to tell the best engineers who have been around for a while that they get to work on something new.
The article also talks about the opportunities ahead.
Tutorial: How to make and share custom GPTs
This tutorial by Charlie Guo explains how to create and share custom GPTs (Generative Pre-Trained Transformers). GPTs are pre-packaged versions of ChatGPT with customizations and additional features. They can be used for various purposes, such as creative writing, coloring book generation, negotiation, and recipe building.
GPTs are different from plugins in that they offer more capabilities and can be chosen at the start of a conversation. The GPT Store, similar to an app store, will soon be launched by OpenAI, allowing users to browse and save publicly available GPTs. The tutorial provides step-by-step instructions on building a GPT and publishing it.
Startup productivity in the age of AI: automate, deprecate, delegate (A.D.D.)
The article by jason@calacanis.com discusses the importance of implementing the A.D.D. framework (automate, deprecate, delegate) in startups to increase productivity in the age of AI. It emphasizes the need to automate tasks that can be done with software, deprecate tasks that have little impact, and delegate tasks to lower-salaried individuals.
The article also highlights the importance of embracing the automation and delegation of work, as it allows for higher-level and more meaningful work to be done. The A.D.D. framework is outlined with steps on how to implement it effectively. The article concludes by emphasizing the significance of this framework in the current startup landscape.
Practical Tips for Finetuning LLMs Using LoRA (Low-Rank Adaptation)
LoRA is among the most widely used and effective techniques for efficiently training custom LLMs. For those interested in open-source LLMs, it's an essential technique worth familiarizing oneself with.
In this insightful article, Sebastian Raschka, PhD discusses the primary lessons derived from his experiments. Additionally, he addresses some of the frequently asked questions related to the topic. If you are interested in finetuning custom LLMs, these insights will save you some time in "the long run" (no pun intended).
The interface era of AI
In this article, the author Nathan Lambert explains the era of AI interfaces, where evaluation is about the collective abilities of AI models tested in real open-ended use. Vibes-based evaluations and secret prompts are becoming popular among researchers to assess models. Deploying and interaction with models are crucial steps in the workflow, and engineering prowess is essential for successful research.
Chat-based AI interfaces are gaining prominence over search, and they may even integrate product recommendations into model tuning. The future will see AI-powered hardware devices, such as smart glasses and AI pins, that will revolutionize interactions with AI. Apple's AirPods with cameras could be a game-changer in this space.
We need your help!
We are working on a Gen AI survey and would love your input.
It takes just 2 minutes.
The survey insights will help us both.
And hey, you might also win a $100 Amazon gift card!
Every response counts. Thanks in advance!
“Math is hard” — if you are an LLM – and why that matters
This article by Gary Marcus discusses the limitations of language models when it comes to solving mathematical problems, particularly multiplication. The author highlights a paper that claims LLMs can solve mathematical problems without a calculator but argues that the results are not as impressive as they seem.
The author compares the performance of LLMs to that of a calculator and concludes that LLMs struggle to generalize and truly understand multiplication. The article suggests that a hybrid approach may be more effective but emphasizes the need for further improvement in LLMs' mathematical abilities.
OpenAI's alignment problem
Even with Sam Altman reinstated as OpenAI’s CEO, we shouldn’t dismiss the possibility that stranger things may still be coming.
OpenAI’s board failed the institution. But did it have a point?
In this interesting post, Casey Newton scrutinizes the board's stated reasons for the firing, which are varied. He also explores the board's role in overseeing OpenAI's mission and tensions arising from the organization's structure, which blends non-profit governance with for-profit incentives. And much more.
In Praise of Boring AI
The author Ethan Mollick praises the boring aspects of AI, highlighting how it can automate repetitive and tedious tasks, making work more efficient and allowing humans to focus on more important things. They share personal examples of how they use AI to quickly complete tasks like creating financial models and finding images for presentations.
Boredom at work is seen as a problem, and AI can help eliminate the unnecessary and repetitive work that leads to boredom. The author suggests that companies should make boring processes "AI friendly" and reward workers for using AI to streamline operations and make everyone happier.
How to create consistent characters in Midjourney
This amazing tutorial by Linus Ekenstam explains how to create consistent characters in Midjourney, an AI diffusion model. It introduces the concept of "image prompting," where images are used as inspiration in prompts to generate desired outputs. The tutorial suggests using human portraits as examples and compares the technique to how illustrators create character blueprints.
It also mentions the importance of good-quality training data for Stable Diffusion models. While this technique can help achieve consistency, it is recommended to use a custom model for Stable Diffusion for better results. The tutorial concludes by mentioning the possibility of future advancements in creating consistent characters and much more.
The Mobile Revolution vs. The AI Revolution
How will AI stack up to past technology revolutions?
This article by Rex Woodbury provides a thought-provoking perspective on the ongoing AI revolution, comparing it to previous technological shifts and offering insights into what the future might hold in terms of innovation and transformation in AI.
The internet, mobile, and cloud looked like their own distinct revolutions– but rather, they may have been sub-revolutions in the broader Information Age that’s dominated the last 50 years of capitalism.
AI is bigger, a more fundamental shift in technology’s evolution.
That's all for now!
Subscribe to The AI Edge and gain exclusive access to content enjoyed by professionals from Moody’s, Vonage, Voya, WEHI, Cox, INSEAD, and other esteemed organizations.
Thanks for reading, and see you tomorrow. 😊