AI Highlights

1. OpenAI's o1: the AI that “thinks” before it speaks.

The wait is over. OpenAI launched o1, an AI model that produces internal chain-of-thought reasoning before responding. It excels in math, coding, and science problems, scoring in the 89th percentile on competitive programming questions and matches PhD-level performance on physics, chemistry, and biology benchmarks.

Alex’s take: We're entering a new era of AI reasoning. I’ve been testing it out this week and can already see a real step-change in complex thought compared to GPT-4o. This could revolutionise fields from scientific research to problem-solving in ways we've only imagined until now.

2. Apple Intelligence is AI for all.

iPhone 16 users enter the AI era with Apple's Intelligence suite. The new features include AI-powered writing tools, video generation from text prompts, smart camera controls, and ChatGPT integration.

Alex's take: Apple is making AI accessible to everyone. By integrating AI directly into the iPhone, Apple is democratising access to these powerful tools. Whilst we're still waiting for access in Europe, there’s no more navigating confusing AI apps or learning new platforms—it's all right there in your pocket.

3. Google's NotebookLM speaks up.

Google has introduced an “Audio Overview” feature to its AI note-taking app NotebookLM. This allows the app to explain complex topics out loud, using AI-generated virtual hosts with conversational speech patterns.

Alex’s take: I can see this being incredibly useful for students, researchers, and professionals who need to quickly make sense of complex information. The ability to download and listen on the go is particularly exciting to turn your commute into a personalised lecture on any topic you're studying.

Content I Enjoyed

AI and the Future of Us.

When Oprah dedicates a primetime special to artificial intelligence, you know it's gone mainstream. I typically avoid talk shows, but (for obvious reasons) I made an exception.

The “AI and the Future of Us” special turned out to be a thoughtful conversation.

Featuring tech heavyweights like OpenAI's Sam Altman and Microsoft's Bill Gates alongside AI critics, it explored AI's societal impact to help viewers navigate this rapidly evolving landscape.

Something that stood out to me was “52% of Americans are concerned about AI”.

Given the amount of fear-mongering the mainstream media pushes (to generate clicks), I thought this was a reasonable statistic.

On balance, it’s also important to consider McKinsey’s estimate that 30% of the global workforce will lose their job to AI within 7 years.

The takeaway? As Bill Gates noted, AI could supercharge fields like education and medicine. But as FBI Director Christopher Wray cautioned, it's “incumbent on everyone in America” to approach AI with both excitement and vigilance. The race is on to harness AI's benefits safely whilst mitigating our downside with bad actors.

Idea I Learned

Human drivers are to blame for most serious Waymo collisions.

Timothy Lee writes “Understanding AI”, a Substack publication about how AI works.

He wrote a great piece on self-driving cars—how they’re significantly safer than human drivers.

Recent data from Waymo, a leading autonomous vehicle company, suggests that their driverless cars are involved in far fewer accidents than human-driven vehicles:

• Injury crashes: Waymo vehicles experience about 70% fewer injury-causing crashes per mile compared to typical human drivers in San Francisco and Phoenix.

• Airbag-deploying crashes: Driverless Waymos are about one-sixth as likely as human drivers to be in crashes severe enough to trigger an airbag.

• Nature of crashes: Out of 23 serious Waymo crashes, 16 involved a human driver rear-ending a Waymo, and 3 others involved a human-driven car running a red light. There were no serious crashes where a Waymo ran a red light or rear-ended another car.

I feel society’s move to widespread autonomous vehicle adoption will be gradual and complex. But this data provides a promising glimpse into a potentially safer future on our roads. I personally can’t wait to have the confidence to take a nap on my next drive.

Quote to Share

Andrew Curran on AI advancement:

“Every major lab has been saying openly for the last six months that we are not at the top of the S-curve, and that they can see nothing but air above us. As it turns out, they were telling the truth. The question now is, how much further can we go?”

An S-curve represents the typical path of progress over time. It's named for its shape, which looks like an elongated “S” when on a graph and has got has three main phases:

1. Initial slow growth (bottom of the S)

2. Rapid acceleration and improvement (middle of the S)

3. Plateau or saturation (top of the S)

The upper limits of AI capabilities are not yet visible. While we know we're not at the top of the curve, we don't yet know how high that curve might ultimately reach or what the full implications of continued AI acceleration might be.

Source: Andrew Curran on X

Question to Ponder

The ultimate goal of integrating AI into our society extends far beyond mere convenience or efficiency—it's about propelling humanity towards becoming a more advanced civilization.

Specifically, our aim is to achieve the status of a Type-1 civilisation on the Kardashev Scale, which requires us to harness all available energy on our planet.

AI stands as our most promising tool to realise this objective.

I thought this video about the Kardashev Scale was a brilliant explanation of what this means for humanity.

💡 If you enjoyed this issue, share it with a friend.