114 | 😇😇👺 Two Truths and a Lie

In today’s Brainyacts:

1. Design failure

2. Google launches NotebookLM

3. Top Brainyacts referrer shoutout!

4. Comprehensive coverage of leading AI model updates

5. News you can use (links to click if the headline grabs you)

6. Awesome new AI app :)

👋 to new subscribers!

To read previous editions, click here.

Lead Story

😇😇👺 Two Truths and a Lie: Generative AI and the Critical Need for Responsible Design

The children's game "Two Truths and a Lie" has found an unexpected parallel in the realm of generative AI. Participants in this game voice two truths and one lie, challenging others to identify the falsehood. Just like the players who must determine what's real or false, users of generative AI face a similar dilemma. Yet, unlike the innocent game, the stakes are much higher, and the consequences are far more significant.

The widespread misuse of AI tools isn't just a failure of understanding on the part of the users; it's a failure of design and responsible communication by those who create and market these tools.

The Flaw of Design: Generative AI's Misleading Game

Generative AI's powerful ability to construct convincing narratives makes for a seductive user experience in many ways. This remarkable capability, however, becomes a double-edged sword without proper guidance and ethical design.

Many AI tools are being unleashed into the market with little to no user manuals, leaving users unprepared for their complexities and potential pitfalls. Poor design choices, such as the absence of disclaimers about the potential for inaccurate information or burying such disclaimers in tiny, unreadable print, contribute to misconceptions and misuse.

Check these screen grabs I just did by the four leading free consumer tools:

OpenAI has a disclaimer but it is small and not highlighted.

Google Bard has a disclaimer but it is small at the bottom of the screen and not highlighted.

Bing Chat says that it does when I asked it as I could not find it. Yet, I still could not find it anywhere in the chat window. Did it hallucinate this!?

Anthropic’s Claude2 has nothing.

💥 These design flaws are not merely oversights; they are a fundamental failure to recognize the social responsibility that comes with powerful technology.

Critical Questions: Addressing Design and Expectation

To navigate the treacherous waters of AI misuse, we must ask and answer several key questions:

• Accountability and Transparency: How can developers ensure that AI tools are transparent in their limitations, clearly communicating the potential for inaccuracies?

• Ethical Design Standards: What standards must be implemented to ensure that AI is designed with user comprehension and ethical considerations in mind?

• Consumer Protections: How should laws protect consumers from being misled by AI tools that lack proper disclaimers and guidelines?

• Special Legislation and Regulation: Is there a need for industry-wide agreements, governmental regulations, or even international collaboration to oversee responsible AI design?

Conclusion: From Misuse to Enlightened Engagement

Generative AI's game of "Two Truths and a Lie" need not be a perilous one. By acknowledging the critical role of responsible design and transparent communication, we can transform these tools from misleading puzzles into empowering technologies.

The future of generative AI must be built on ethical design principles that prioritize clarity, transparency, and user education. Disclaimers must be clear and accessible, not hidden in fine print. User manuals must be comprehensive and readily available, guiding users to engage with the technology responsibly.

The lessons learned from the simple game of "Two Truths and a Lie" resonate profoundly in our interaction with AI. We must approach AI not as mere players in a game of chance but as informed participants in a well-designed system that respects our intelligence and agency. Only then can we fully embrace the potential of AI while upholding the principles of truth, integrity, and ethical engagement.

Spotlight Story

Google just released a new Personal AI tool for your Google Drive/Docs.

Here is a short video of my first test drive.

Here is a Twitter post with some other details.

Top Brainyacts Referrer Shout Out!

Since the beginning of this newsletter, I have made it a point to celebrate and showcase, you, the reader. For those that think highly of Brainyacts and share it with others, I thank you. BTW everyone has a unique link to share at the bottom of every newsletter). For those that drive 10+ new subscribers, you get a special shout-out.

Thank you to Lee Rosen! Lee has been a reader for over 100 editions. And I love his story and am totally amped to have readers like him.

Learn about Lee in his own words:

AI Model Notables

• OpenAI rolls out some updates:

1. Prompt examples: A blank page can be intimidating. At the beginning of a new chat, you’ll now see examples to help you get started.

2. Suggested replies: Go deeper with a click. ChatGPT now suggests relevant ways to continue your conversation.

3. GPT-4 by default, finally: When starting a new chat as a Plus user, ChatGPT will remember your previously selected model — no more defaulting back to GPT-3.5.

4. Upload multiple files: You can now ask ChatGPT to analyze data and generate insights across multiple files. This is available with the Code Interpreter beta for all Plus users.

5. Stay logged in: You’ll no longer be logged out every 2 weeks! When you do need to log in, you’ll be greeted with a much more welcoming page.

6. Keyboard shortcuts: Work faster with shortcuts, like ⌘ (Ctrl) + Shift + ; to copy last code block. Try ⌘ (Ctrl) + / to see the complete list.

• OpenAI Plugin store now 800 plugins (IMO many are junk! - use caution)

• Google ‘mind-reading’ AI can tell what you’re hearing by monitoring your brain signals in eerie experiment

• 3 new things you can do with generative AI in Google Search

• Generative AI for audio made simple and available to all via Meta

News You Can Use:

➭ Water usage soaring by cloud providers due in part to AI 💦

➭ Legal services pricing specialists show back of the napkin impact of Generative AI on law firm profitability

➭ Dentons to launch its own ChatGPT-like model

➭ 7 draft principles on responsible use of AI in legal - seeking public comment

➭ The AI Rules US policymakers are considering

➭ AI in the regulatory tech (regtech) arena: what to ask

➭ How hedge fund managers are using ChatGPT

➭ New McKinsey detailed and in-depth Generative AI report

➭ How Warren Buffett thinks of AI

➭ The GenAI tug-o-war between companies and hackers has begun

➭ A list of “best” AI legal assistant tools

➭ LexisNexis Brings Lexis Create to U.S., Announces New Gen AI Capabilities Ahead