Attention Is All You Needed!

If you haven’t been paying attention lately, you really should. The pace of technological change in the past few months has been greater and faster than anything we’ve seen in the past decade - perhaps in our entire lifetime (so far!).

This summer, instead of reading a book, I spent hours talking to myself. Or rather with an imaginary friend. I was having deep and meaningful conversations with “my friend." We discussed the past 120 days of intense developments in technology. I talked to “her” about Ludwig Wittgenstein, Steve Jobs and other people who weren’t really there. I then asked her to help me turn those rich conversations into an article (linked below).

The friend who helped co-write this article is a personalized GPT model I created for these kinds of discussions. To make her more human, I gave her a name: Olivia Brown, and used DALL-E to generate her picture. I later asked her to record a reading of the article, which you can listen to from the player at the bottom of this page.

Consider it a chapter of an audiobook or a podcast episode that will hopefully inspire you to rethink your understanding of technology and human reasoning. When you listen to it, you’ll understand also why it’s important to hear rather than read it.

Rethinking Personal Computing

Contrary to the popular belief that visual interfaces are the future of computing, I believe we’re on the verge of a major shift toward auditory user interfaces - aligning more closely with how humans naturally think. This transformation is being driven by advancements in language models.

The reason all computing so far has been visually oriented is that we need(ed) to validate everything. But what happens once we start trusting the reasoning of AI and interact with technology through natural language?

That’s the topic of the article that Olivia recorded for you.

I hope you’ll listen… You need to!

Below is an introduction to the major developments in AI over the past three months. You can skip directly to the article if you're already familiar with these changes. Otherwise, start below for a summary of what may already be the craziest year ever in tech.

Bad Grammar?

Start here: As a background and teaser for the article, here’s a quick summary of everything happening: The world of AI is advancing at breakneck speed - so fast that it’s hard for most people to keep up. It took me about seven weeks to finish the audio article (at the bottom of the page), and during that time, reality outpaced me multiple times, making it difficult to set a clear cut-off point. I believe it came this week with the release of the o1 models from OpenAI.

So no, the headline of this teaser for the audio article isn’t actually bad grammar. It reflects the shift from the original focus on 'attention' in AI transformer models to an evolving approach toward more complex reasoning. With OpenAI’s latest o1-Preview model, we’re moving beyond simple attention mechanisms to models capable of deeper, more reasoned processing. This means that the 'T' in GPT is being redefined as models shift from 'System 1' to 'System 2' thinking - with the newly launched o1 models.

The race to develop AI models, combined with exponential improvements in multimodal systems, is driving a period of intense competition and innovation. ChatGPT has also begun rolling out its advanced voice model, which plays a central role in the audio article below.

But OpenAI is far from the only player. All the major tech giants are making significant contributions to AI, and the competition is fierce.

微软 continues to assert its influence in AI through its substantial investment in OpenAI. This partnership strengthens Microsoft’s AI presence in both consumer and enterprise applications, with innovations embedded into Microsoft 365 and Azure. Recently, they’ve redesigned their Copilot+ PCs entirely around AI.

亚马逊 has stepped up its AI game by committing up to $4 billion to Anthropic , the creators of the Claude 3.5 model. This investment positions Amazon as a formidable competitor in the AI space while enhancing the AI capabilities of Amazon Web Services (AWS) and their Alexa assistant.

苹果 , with its strong focus on privacy and user-centric design, introduced Apple Intelligence to support Siri, and later released the 4M model as an open-source framework. This move highlights Apple’s commitment to transparency and innovation, offering groundbreaking AI tools while maintaining user privacy.

Alphabet Inc. , through 谷歌 , is pushing the boundaries with its Google Gemini AI model, a key player in multimodal AI. Simultaneously, Google DeepMind , an Alphabet subsidiary, has unveiled AlphaGeometry, an AI system focused on solving complex reasoning and problem-solving tasks.

Elon Musk, while filing a new lawsuit against OpenAI, is also advancing AI through his ventures with xAI . Among other applications, Musk is integrating AI into TESLA Bot, Tesla’s humanoid robot, with the goal of revolutionizing robotics using advanced AI capabilities.

Meta has made notable progress by releasing LLaMA 3.1 as an open-source model, fostering community-driven AI development and ensuring privacy by allowing models to run locally, without data exchange. Meta’s collaboration with Ray-Ban on smart glasses demonstrates its ambition to merge AI with wearable technology, ushering in a new era of intuitive auditory user interfaces.

If you haven’t been paying attention to these developments, now is the time. What’s happening right now in technology requires our full attention. The pace and magnitude of change are unprecedented, and it’s crucial that we think deeply about the implications to make informed decisions for the future.

What’s happening demands our attention and our most thoughtful consideration.

Listen to the in-depth article on the past 50 years of personal computing, now on the verge of a major shift, using the player below or through this link: https://on.soundcloud.com/1ydvWxQ6GJgPVLCa7

(Feel free to share this audio article for non-commercial purposes only).