8 Google employees invented modern AI. Here’s the inside story
Illustration: Magda Antoniuk

8 Google employees invented modern AI. Here’s the inside story

Hello, and welcome to WIRED Start: your weekly roundup of the most important stories, landing in your inbox every Monday. Don’t forget, you can get the very latest from WIRED with our daily newsletter. Sign up to receive it for free here.


There are eight names listed as authors on "Attention Is All You Need," a scientific paper written in the spring of 2017. They were all Google researchers. Those eight researchers couldn't have known when they sent their paper off to a prestigious artificial intelligence conference that it would kick off a worldwide AI revolution.

This week, WIRED Start looks into the story of the eight Google employees that invented modern AI. Approaching its seventh anniversary, the “Attention” paper has attained legendary status. The authors started with a thriving and improving technology—a variety of AI called neural networks—and made it into something else: a digital system so powerful that its output can feel like the product of an alien intelligence. Called transformers, this architecture is the not-so-secret sauce behind all those mind-blowing AI products, including ChatGPT and graphic generators such as Dall-E and Midjourney.

All eight authors have since left Google. Like millions of others, they are now working in some way with systems powered by what they created in 2017.?

Steven Levy talked to the Transformer Eight to piece together the anatomy of a breakthrough, a gathering of human minds to create a machine that might well save the last word for itself.

Read the full story here.


Get ahead with these recommended reads

Hackers Found a Way to Open Any of 3 Million Hotel Keycard Locks in Seconds

Photo-Illustration: WIRED Staff; Getty Images

The company behind the Saflok-brand door locks is offering a fix, but it may take months or years to reach some hotels, writes Andy Greenberg .

Read the full story here.


Here’s Proof You Can Train an AI Model Without Slurping Copyrighted Content

Illustration: Jacqui Vanliew; Getty Images

OpenAI claimed it’s “impossible” to build good AI models without using copyrighted data. An “ethically created” large language model and a giant AI dataset of public domain text suggest otherwise, writes Kate Knibbs .

Read the full story here.


Watch Neuralink’s First Human Subject Demonstrate His Brain-Computer Interface

Courtesy of Neuralink

In a livestream on X, the paralyzed 29-year-old man used his Neuralink brain implant to control a computer, writes Emily Mullin .

Read the full story here.


The Apple Antitrust Case and the ‘Stigma’ of the Green Bubble

Photograph: Justin Sullivan/Getty Images

The US government's Apple lawsuit leans on the social cost of not owning an iPhone, an unusual argument for antitrust, writes Lauren Goode .

Read the full story here.


Until next time

Illustration: Jared Bartman; Getty Images

Technology and politics have never been more intertwined. On a global scale, 2024 is poised to be the most purely digital, technologically mediated, psychologically manipulative election season any of us have ever seen.?

Across the US election and many more—including key votes in India, Ukraine, and Mexico—we’re going to examine both the campaigns and the election stakes through a digital lens, focusing on disinformation, online extremism, conspiracies, social platforms, election influencers and grifters, the influence of AI, and more.?

Let Makena Kelly and the WIRED Politics team guide you through the vortex of extremism, conspiracies, and disinformation. Sign up for the weekly newsletter here.


Thank you for reading! We'll be back next Monday with another WIRED Start.?

If you enjoyed reading this week’s edition, you can unlock unlimited access to WIRED’s content with a subscription.

Zerhio Zenil

Marketing Manager

8 个月

Interesting publication; however, it's important to mention that there is already a new paradigm: smaller models instead of colossal training sets. In other words, the era of big data is over, and the days of ever-larger models, the 'L' and the 'M' in 'LLM', are numbered, mainly because we have exhausted them. Here is the opinion of Oxford researcher Dr. Hector Zenil regarding LLMs. Follow him → https://shorturl.at/bilCW

Michael Owens

Providing Factory 4.0 traceability solutions with decades of experience and an open mind... Socrates says, “The only true wisdom consists in knowing that you know nothing.”

8 个月

Google AI is all blue biased & no red according to an article I read recently. All the AI can be tested for designer bias. It will be interesting to see how the sheeple conform,. Make it easier for us all. Make it quick. ?? ??

Rosemary Hood

Rosemary Hood DVM Emerita

8 个月

Complete lack of governance .... jurisdiction ... ?? but that was known, so what! WAKE UP

回复
Zakaria Khan

Business Owner at TKT home made mosla products

8 个月

Great share WIRED

??? #NewCareerThisYear2024 ??

回复

要查看或添加评论,请登录

WIRED的更多文章

社区洞察

其他会员也浏览了