The Open-Source War Begins

You can mail a coconut without packaging. In the U.S., it’s perfectly legal to send an unwrapped coconut through the mail. Just slap on some stamps and an address—no box needed! We need to try this.

What’s in store:

• Open-sourcing AI is a lot more complicated than you think.

• Why China didn’t sign the military AI agreement — let’s find out!

• Around the web: Google is testing a new Android tablet feature allowing users to resize apps like desktop windows.

• Why we Guidde for all of our documentation!

Read Time: 6 minutes

AI POLICY

The open-source AI war has begun

The Open Source Initiative (OSI) is getting closer to agreeing on a crucial definition of open sourcing when it comes to AI systems.

The announcement is expected in October, but some open-source champions are already distancing themselves from it.

This is because of disagreements about how the systems operate, with a three-level system proposed. Yeah, there’s actually beef between the open-source gurus.

The argument

Jim Zemlin detailed The MOF - Model Openness Framework - which he states is a way to help evaluate if a model is open or not open.

Within the MOF, there are three tiers of openness. The highest level, level one, is an open science definition where the data, every component used, and all of the instructions must go and create your model the same way.

Level two is a subset where not everything is open, but most are. Then, on level three, you have areas where the data may not be available, and the data that describe the data sets would be available.

The problem? Disagreement with this rating system.

Tara Tarakiyee, FOSS Technologist for the Sovereign Tech Fund, says that any system that must use certain data cannot be open source.

This is echoed by open-source leader Julia Ferraioli, who says the definition risks diluting what open source truly means.

Here’s what you should know:

• The OSI is close to agreeing on a definition of open-source AI.

• Leading figures in the open-source world are not happy with the definition.

• The argument risks damaging the OSI and leaving unclear guidelines on AI.

What about Meta?

Meta already claims their LLama model is open source, but it isn’t really.

That’s because it doesn’t use an OSI license. Meta forbids the use of the model to train other AI models for example - any kind of restriction means it can’t truly be called open source - for now at least.

So, it’s all a bit more complicated than it would first seem.

Who knew open-sourcing was so complex?

Vote for live results and see results + opinions from yesterday at the bottom of the email.

PRODUCTIVITY HACKS

Explain it once, and let AI handle the rest

Ever wish you could clone yourself to answer the neverending repetitive questions?

Well, with Guidde, you can!

Guidde is an AI-powered tool that can turn your instructions and explanations into neat, easy-to-follow videos in seconds.

It’s like having a digital assistant who never gets tired of explaining the same old thing for the nth time.

Whether it’s showing your team how to navigate a new software or helping customers get the most out of your product, Guidde makes it happen in record time.

Here’s how Guidde gets it done:

• Turn boring explanations into dynamic video tutorials with just a few clicks

• Keep your team on the same page with branded, consistent content

• Update and edit videos with brand-new features like intros and outros

Whether you’re training new hires, or keeping your customers in the loop, Guidde does the hard work so you can focus on other things.

Trusted by 20,000+ users and rated 5 stars on the Google Chrome store, it’s the go-to tool everyone should have.

Why not give it a try? It’s free!

On This Day

September 17th 1787 - The United States Constitution is signed in Philadelphia, Pennsylvania, by delegates to the Constitutional Convention.

Weird AI

The weirdest pizza ad in history

AI tried its hand at creating an old-school pizza ad, and the result? Well, it's weird. The ad for the fictional "Pepperoni Hug Spot," made by a Twitter user exploring AI, definitely brings some nostalgia for those classic American pizza commercials—but it also feels totally off.

From awkward humanoids trying to use their lips to the strange, jumpy flow of the video, it's clear AI still doesn't quite get how humans do things (like, you know, eat). Still, with taglines like "like family, but with more cheese," it’s not all bad—just another step in the weird world of AI experiments.

AI POLICY

China’s reluctance to sign the military AI deal

China recently decided not to sign an agreement aimed at limiting the use of AI in military settings.

This happened during the Responsible Artificial Intelligence in the Military Domain (REAIM) summit in South Korea.

About 90 countries were represented, with around 60 nations, including the U.S., agreeing to a nonbinding plan.

This plan focuses on keeping human control in military AI systems.

However, China, along with 30 other countries, chose not to agree.

Arthur Herman, an expert on AI from the Hudson Institute, explained that China’s choice is typical of its approach to international agreements.

Beijing often avoids treaties it hasn't helped design, viewing them as potential constraints on its military AI development.

Herman said that China is cautious about such agreements because they could restrict its ability to advance its technology.

Here are some key findings:

• The agreement stresses that human control in military AI decisions is vital.

• China, along with 30 other nations, did not sign the agreement.

• The plan represents a global effort to introduce ethical oversight in military AI.

Thanks, but no thanks

The plan also aims to deal with concerns about how quickly AI systems can make decisions, especially in life-or-death situations like using weapons.

Countries like the U.S. believe that human oversight is necessary to ensure that important decisions remain in human hands.

Even though China didn’t sign the agreement, the blueprint is still a significant effort by many nations to promote ethical standards in military AI technology.

Ethical AI? Yeah, that’s kind of important when we’re talking about missiles.

Mindstream Picks

Space: A Crew Dragon spacecraft splashed down in the Gulf of Mexico on Sept. 15, ending a five-day private astronaut mission that included the first commercial spacewalk.

Crypto: Shiba Inu opened at $0.000014 on Sept. 15, its highest in 20 days, while bear traders set a sell-wall to block a breakout above $0.000015.

Games: Italian police seized €50m worth of fake vintage consoles and games, including Mario Bros., Street Fighter, and Star Wars.

Music: The Cure may have confirmed the release date of their first LP in 16 years via postcards sent to fans.

Don’t Miss - Google is testing a new "desktop windowing" feature for Android tablets, allowing users to resize apps and arrange them freely, similar to iPad's Stage Manager. Available as a developer preview, it supports multiple app instances and includes a taskbar for quick access to open apps.

AI Art

Our Image of the Day

Daily Image Prompt

Submit your artwork to Mindstream →

Yesterday’s Poll

“How do you feel about Meta using public posts to train its AI?”

It's a smart move for AI development. 34%

It crosses the line when it comes to privacy. 66% ✅

Your Views:

“I'm unhappy about it and wish to opt out. How do they deal with posts which include children under 18?” - julie@emerald

“It opens to many doors; where is the line drawn, who is responsible for maintaining its position, if moved once why not twice, three or more times?” - jeffreykey

Submit your opinions in our daily poll to be featured!