How AI Works, In Plain English

Brought to you by:

The DPAA is using environmental DNA in the Great Lakes to trace human remains on shipwrecks, leveraging the region's optimal preservation conditions to aid global archaeological research.

Welcome, AI and tech fans!

Here’s what’s happening today:

  1. OpenAI’s Sam Altman backs teens’ startup

  2. Artifact now lets users generate AI images for posts.

  3. How AI works, in plain English.

  4. Arc Browser introduces unique AI-powered features.

1. Sam Altman Backs Teens’ Startup

Sam Altman, along with other notable backers, is supporting a teen-founded AI startup, Induced AI, which aims to revolutionize business workflow automation.

Induced AI allows businesses to describe their workflows in plain English, which the platform then translates into pseudo-code for various repetitive tasks.

The platform operates Chromium-based browser instances, enabling interactions with websites even without an API. Unlike traditional methods that require manual tagging of HTML elements, Induced AI can interpret instructions from English and adjust them dynamically.

The startup recently secured $2.3 million in seed funding and is distinguishing itself in the Robotic Process Automation space by offering features like simultaneous task execution and full remote capabilities. TechCrunch

2. Artifact, News App Turned X Competitor, Allows Users to Generate Images

Artifact, originally a news aggregator, has introduced a generative AI feature allowing users to create images for their posts, enhancing storytelling and engagement.

Developed by Instagram's co-founders, Artifact has evolved from solely news aggregation to a platform with social features like commenting and user profiles, positioning itself as a competitor to platforms like X.

The new AI-driven image generation, using the Stable Diffusion model, offers users quick and customizable visuals to accompany their content, further personalizing user experience. TechCrunch

3. How AI Works, In Plain English

Scott Rosenberg of Axios recommends three comprehensive articles for understanding how generative AI, like ChatGPT, functions. The articles range from beginner to advanced levels:

  1. "How transformers work" from The Financial Times explains the "transformers" approach to neural network design introduced by Google researchers in 2017.

  2. "Go deeper into large language models" from Ars Technica delves into how large language models predict the next word in a sequence using "word vectors" and machine-learning "neurons."

  3. "The mathematician's perspective, for the rest of us" by Stephen Wolfram offers a deep dive into how AI programs translate everything into numbers, discussing the vast scale and the "temperature" settings influencing AI responses.

While AI may seem intricate at first glance, with a bit of effort and the right resources, its complexities can be unraveled for anyone eager to understand.

4. “Arc Max” Released As AI-Powered Browser

The Arc browser branded as "Arc Max," integrates with OpenAI's GPT-3.5 and Anthropic's models to craft AI-powered functionalities. While it has capabilities akin to other AI assistants in competing browsers, like engaging with ChatGPT, Arc Max offers unique enhancements:

  • It can rename pinned tabs using the page title, making them concise and readable.

  • It can rename downloaded files based on their content.

  • It provides a summary preview of links when users hover over them and press shift.

Earlier, Arc had released a "Boosts" feature to customize web page elements. The company explored allowing users to create Boosts via prompts, but these didn't make the final cut due to speed concerns.

CEO Josh Miller, in a livestream, mentioned that the new features are subject to change based on user feedback over the next 90 days. TechCrunch

More Notable Headlines

Meta and other major social media platforms are introducing or testing premium subscription tiers, marking a shift from the traditional ad-supported model to one where users pay for enhanced experiences, potentially leading to a decline in user experience for non-paying users, similar to the airline industry's model of charging for previously free amenities.

Elon Musk advised publishers to share content in long form on X (formerly Twitter) as the platform's algorithm prioritizes time spent on X, implying that links receive less attention; this comes amid criticisms of Musk's content moderation changes and concerns about misinformation on the platform.

That’s it for today!

We’re new and eager for feedback. Tell us how we can make the Great Lakes Beacon more relevant to you! Simply reply to this email with your comments and suggestions. 🤗