072 | The Dark Web

OpenAI releases the official ChatGPT app for Apple devices (Android coming soon)

Let’s keep learnin’.

A special welcome 👋 to my NEW SUBSCRIBERS! 

To read previous posts, go here.

In this edition we will

1. ply the depths of the Dark Web

2. enable and use ChatGPT plus plugins

3. use some new AI jargon

4. see how today’s Supreme Court decision impacts AI regulation

5. share an upcoming AI Ethics free online event

Accessing the Dark Web: A Crime Fighting AI Hero Emerges

🚇The dark web is a part of the internet that isn't visible to search engines like Google. It's like an underground network that you can't access with a regular browser or without special tools.

🧊Think of the internet as an iceberg. The part of the internet we use every day (like reading news, shopping, or social media) is just the tip of the iceberg that's above water. This is often called the "surface web."

🔦Below the water is the rest of the iceberg, much larger and hidden. This is where the dark web exists. It's a place where people can communicate and share things anonymously, which means it can be used for both good and bad purposes.

⚠️🏴‍☠️For example, it can be a safe space for political activists who live in countries where their speech is restricted. But, it's also a place where illegal activities can happen, like selling stolen data or illegal substances.

🔐To access the dark web, people usually use a special browser called Tor, which helps them stay anonymous. But remember, just because you can visit the dark web, doesn't mean you should. It can be risky and potentially illegal, depending on what you do there.

Why am I sharing this with you?

Because the dark web plays a major role in everything from major fraud schemes, malware sharing, terrorist plotting, and thousands of illegal transactions and deal. The rise of crypto has been like adding lighter fluid to an already hyper-active underground.

Authorities around the globe have teams of experts trolling, tracking and monitoring the dark web constantly, but the sheer size and number of tentacles make it an enormous challenge.

Enter generative AI and a new model called DarkBERT.

DarkBERT is a revolutionary AI model trained on the elusive data of the dark web. This South Korean innovation, built on the RoBERTa architecture, an older LLM that was undertrained compared to today’s practices has been dusted off and is being trained once again. Early returns shows that it is set to redefine our understanding of the dark web's coded dialects and clandestine communications.

As AI applications surge, DarkBERT stands out, promising to empower security researchers and law enforcement with unprecedented access to the web's hidden corners. Yet, the journey has just begun, with further enhancements on the horizon. Discover the intriguing story of DarkBERT's creation and its potential implications in our latest feature.

FREE 5-day email course.

🆓 ⭐️ FYI I built this for anyone brand new to ChatGPT

• Beginners Introduction to ChatGPT
• 10X Your Productivity with ChatGPT

Sign up here: https://brainyacts.joshkubicki.com/course

Tool you can Use: ChatGPT Plugins

Some time ago, OpenAI announced that plugins would be coming available to all ChatGPT plus users (paid subscribers).

This along with web-browsing has been quietly rolled out over the last several weeks. This week OpenAI stated that all Plus subscribers would be getting access to plugins.

When you are granted access, you do not get a note or notification of any kind - they just appear as new options. You might miss them if you are not paying attention.

So, today let me walk you through how to:

• get on the waitlist for plugins, if you are not already

• check to see if you have access

• start using plugins if you do have access

Getting on the waitlist:

go to https://openai.com/waitlist/plugins

Just click on Join plugins waitlist to enter your information. You will be asked the following:

“What types of use cases or new plugins do you want to see built?“

This is not required to be answered but I encourage you to as OpenAI is reviewing all of this as it continues to develop and refine its products.

Checking to see if you have access:

Go to your settings - you will find them by clicking on the three dots next your name in the lower left-hand corner of the main ChatGPT screen.

When you click the three dots, the following window will appear.👇 Here you will see in the left-hand side if Beta features have been turned on for your account and which ones you have access to. You will need to switch on (green slider) the ones you want to enable.

If you do have plugins and you turn them on, you won’t see anything change yet until you go back to the main chat window.

🚨Note: You may need to reload the page in order to enable the change.

1. At the top of the page, click on the down arrow next to GPT-4.

2. Click on plugins so the blue check mark appears, you will then see a down arrow appear underneath the GPT-3.5 and GPT-4 slider.

3. Click on that to access the plugin store. See the Gif here and pics below:

🚨‼️Note: Read the About Plugins information that appears at the bottom center of the plug-in store.

Using plugins:

You cannot direct which plugins to use when. It is all based on your prompt. So depending on what you are trying to do and how you ask/prompt, OpenAI may or may not utilize a plugin.

I tested out FiscalNote - which is a global policy and regulatory insight and updates.

Once I installed it I then used the following prompt:

▶︎▶︎PROMPT

What are the latest AI regulatory actions, hearings, news items from around the globe from within the last 24 hours?

It worked to trigger the FiscalNote plugin:

Now notice that it isn’t quite perfect as I asked for the last 24 hours and it went well beyond that. This could be the result of OpenAI misinterpreting the prompt or a limitation in what FiscalNote allows access to, as it is a paid service so it may limit what the plugin can retrieve.

Regardless, using OpenAI to access real time information is a welcomed addition.

News you can Use: 

AI Jargon Alert: AI-Washing

AI-washing is a growing concern in the tech industry, where companies claim to use artificial intelligence (AI) in their products or services, but in reality, they don't, or they exaggerate its capabilities. This practice is similar to greenwashing, where companies falsely or misleadingly promote their products as environmentally friendly.

Sorry to say but in the legal tech landscape, over-hyping, over-promising, and over-promoting technology capabilities is so dominant that it is hard to not be a pure cynic or skeptic about any claim. I truly feel for legitimate legal tech companies out there.

Back to AI-washing. This term came into the spotlight when an Indian startup, Engineer.ai, was caught making false claims about its AI-based app development platform. The company claimed that its "human-assisted AI" technique could build more than 80% of a mobile app from scratch in less than an hour. However, it was later revealed that the "AI" was actually a team of software engineers. That’s a heavy dose of humans, not AI!

AI-washing is not an isolated incident. A study of 2,830 European startups found that 40% of companies describing themselves as 'AI startups' used virtually no AI at all. The reason behind this is that AI is a buzzword that attracts investors and customers. Startups claiming to use AI can attract up to 50% more funding than other software firms.

The US Federal Trade Commission (FTC) has warned companies against making exaggerated claims about their use of AI. AI-washing doesn't only mean falsely claiming to use AI; it can also mean overstating what an AI product can do. Claims are considered inaccurate if they lack scientific support or only apply under certain conditions.

AI-washing is becoming a serious issue, much like greenwashing. Just as there is increasing scrutiny and legal enforcement against greenwashing, we might see the same for AI-washing.

So, how can you detect AI-washing? 

• Look for evidence of AI's application in a company's product or service.

• Be skeptical of vague or exaggerated claims.

• Remember, true AI should be able to handle unstructured data, not just simple algorithms or machine learning. And finally, keep an eye out for regulatory warnings or actions against companies for false AI claims.

Supreme Court Ruling that will impact AI regulation

The U.S. Supreme Court today upheld Section 230, a law that protects social media platforms like Google, Twitter, and Facebook from liability for user-generated content. This decision safeguards these tech giants' current business models but sparks ongoing debates about the law's relevance in today's digital age.

Section 230, enacted in 1996, allows platforms to moderate content without being held responsible for what users post. Critics argue that these platforms have misused their immunity, and changes to the law could drastically alter online communication.

As for AI regulation, this decision sets a precedent that could influence how future laws treat AI-generated content. If platforms aren't held liable for user content, what happens when the "user" is an AI? This could lead to complex questions about responsibility and accountability in the age of AI, potentially shaping the future landscape of AI regulation.

May 23rd Free Online Event about Ethics and AI.

Project Voice Capital Partners is a new venture capital firm identifying and investing in venture-grade early-stage conversational AI and/or voice technology companies.

Project Voice 2023 recently took place in April in Chattanooga, Tennessee, and is the main industry conference for conversational AI and voice technology here in the United States. Over 2500 attendees and 100+ media outlets attended.

DISCLAIMER: None of this is legal advice. This newsletter is strictly educational and is not legal advice or a solicitation to buy or sell any assets or to make any legal decisions. Please /be careful and do your own research.8