248 | 💼 💰 Venture capitalists as DC lobbyists

In today’s Brainyacts:

1. VC goes DC

2. Underwhelmed with GenAI?

3. Lexis+ AI still hallucinating? and other AI model news

4. Gramm-bot fools phone scammers and more news you can use

   👋 to all subscribers!

To read previous editions, click here.

Lead Memo

😏 🏛️ The Rise of Lobbying Arms in Venture Capital: The Case of Andreessen Horowitz

I just learned that Facebook alum Matt Perault is joining Andreessen Horowitz (a16z) as head of AI policy, based out of their D.C. office. That caught my attention because, let’s be honest, venture capital firms diving into lobbying is still a pretty new thing.

Why Are VCs Getting Into Lobbying?

Traditionally, VCs weren’t exactly power players in D.C. They were busy finding and funding the next big thing, leaving the policy battles to industries like energy or healthcare. But now? The game has changed. Technology is shaping every part of our lives, and governments are starting to regulate areas like AI and cryptocurrency with a heavy hand. For firms like a16z, sitting on the sidelines just isn’t an option anymore.

In 2024, a16z spent over $1.39 million on lobbying efforts and even opened a dedicated D.C. office. Hiring someone like Matt Perault—a seasoned policy expert—to focus on AI shows they’re serious about steering the rules that could make or break the industries they invest in. It’s a bold move, but one that’s starting to become more common among top-tier VC firms.

Why a16z’s D.C. Strategy Stands Out

This isn’t just a one-off hire. a16z is embedding itself into the heart of policymaking, focusing on areas like AI and cryptocurrency that are poised for massive regulatory shifts. These are high-stakes arenas:

• AI: Everyone’s racing to regulate it—whether it’s about safety, ethics, or economic impact.

• Crypto: With so much uncertainty around blockchain rules, getting ahead of regulation could determine the future of entire markets.

a16z isn’t just watching the game; they’re stepping onto the field. Most VC firms don’t have the resources or, frankly, the guts to play this big, but a16z sees the long-term benefits of being at the table when policy decisions are made.

What’s Driving This Trend?

So why are VC firms like a16z suddenly flexing their lobbying muscles? A few reasons:

1. Regulations Are Coming Fast: Governments are scrambling to figure out how to handle emerging tech. If firms don’t speak up now, they risk being hit with rules that could stifle innovation—or their profits.

2. Big Tech Dominates Lobbying: Companies like Google and Meta spend huge sums on lobbying, often pushing agendas that don’t align with startups or early-stage innovations. VCs need to step in to level the playing field.

3. Geopolitical Stakes: Technologies like AI are now tied to national security and global competitiveness. VCs are positioning themselves as allies in strengthening U.S. leadership while countering rivals like China.

4. Opportunities in Public-Private Partnerships: With governments investing heavily in areas like AI infrastructure, there’s a chance for startups (and their investors) to benefit—if the policies align.

What Does This Mean for Venture Capital?

This move by a16z signals a bigger shift for the industry. As regulations grow more complex, lobbying is becoming less of a “nice-to-have” and more of a necessity—at least for the big players. Firms with deep pockets, like a16z, are leading the way, but smaller VCs may struggle to keep up.

At the end of the day, this trend highlights something important: innovation doesn’t happen in a vacuum. Whether it’s AI, crypto, or biotech, the future of these technologies depends not just on great ideas but also on smart, forward-thinking policies. And VC firms like a16z are making sure their voice is heard.

Spotlight

👋 Underwhelmed with GenAI? The problem is you!

Here’s the core truth: every single character you type into a prompt matters. The AI doesn’t just “read” your input like a human would; it vectorizes and mathematically correlates each part of your text to determine the most fitting response. Subtle differences in phrasing, word choice, or even punctuation can significantly shape the output.

For example:

• Asking the model to “create” something may yield a different tone, structure, or level of detail than asking it to “make” the same thing.

• Leaving off a period in a sentence may subtly influence how the AI interprets the level of formality you’re aiming for.

• Including context or clarifying intent within your prompt dramatically increases the specificity and usefulness of the response.

Here is a view of how your words “look” to the model as vectors:

Most new users, however, enter overly basic prompts like “Write an email” or “Tell me about X.” These generic commands do work, but they strip the AI of the opportunity to deliver its full potential. Without nuance or direction, the AI can only deliver something equally broad or surface-level. It’s like asking a master chef to make “food” instead of requesting a carefully crafted three-course meal. The chef can cook, yes—but without more direction, you’re leaving the brilliance of their craft untapped.

The Secret to Better Outputs: Be Deliberate with Your Inputs

To truly unlock the power of generative AI, you have to understand this: your input sets the stage for the AI’s output. The model will work hard to interpret your intent—because that’s what it’s designed to do—but the precision and potency of the response are directly proportional to the effort and specificity you put into crafting the prompt.

This doesn’t mean you need to write essays for every prompt. Instead:

• Be intentional. Use words that reflect exactly what you want. Replace vague terms with specific ones.

• Provide context. Add a sentence or two to explain the purpose of your request. Is this a formal email? A witty social post? A technical report?

• Iterate. If the first output isn’t quite right, refine your prompt. Add examples or clarify your desired tone.

Here’s an example of the difference:

Basic Prompt:

“Tell me about AI.”

Refined Prompt:

“Write a conversational explanation of AI for a newsletter audience, focusing on how it transforms everyday tasks. Use an engaging but approachable tone.”

The second version primes the AI with what you want and how you want it, dramatically increasing the relevance and quality of the response.

AI Model Notables

► Law professor gives Lexis+ AI a failing grade. Gives you the prompts and screenshots of the responses.

► Elon Musk escalates his fight with OpenAI and Microsoft, filing a revised lawsuit accusing them of monopolistic practices and anti-competitive collaboration in the generative AI sector. Read Amended Complaint HERE.

►OpenAI’s upcoming AI agent “Operator,” expected in January 2025, aims to transform task automation and digital interaction, offering features like web browsing, form filling, and personalized assistance. Operator faces competition from AI tools like Anthropic’s “Computer Use” and Google’s developing agent, but its potential lies in its general-purpose capabilities for smoother, more efficient workflows.

► Also, ChatGPT for macOS can now integrate with select apps like VS Code, Xcode, and TextEdit, allowing users to collaborate with the AI without copy-pasting text. This “Work with Apps” feature is a foundational step toward AI agents that seamlessly interact with all apps, making ChatGPT more helpful for both technical and non-technical users—part of OpenAI’s broader move toward general-purpose AI tools like the upcoming “Operator” shared above.

► ChatGPT’s Windows app is now available for everyone.The app launched last month, but was initially was only available for paid users. Now, all users can try it, OpenAI says.

► Google has brought its Gemini artificial intelligence (AI) offering to the iPhone - users can access Google’s Gemini Live voice assistant as well as its premium image generator, Quartz.

News You Can Use:

➭ This is hilarious! UK telecom creates “Daisy” chatbot that flips the script on phone scammers, wasting scammers’ time with fake elderly banter while keeping real victims out of harm’s way.

➭ Nearly half of AI data centers may not have enough power by 2027.

➭ Apple’s rumored six-inch ‘AI wall tablet’ could control your smart home by March 2025.

➭ Post-Election outlook for regulation of AI + Trump expected to trash Biden’s AI agenda.

➭ Google Street View but for forests and caves? This AI startup is doing it.

Who is the author, Josh Kubicki?

Some of you know me. Others do not. Here is a short intro. I am a lawyer, entrepreneur, and teacher. I have transformed legal practices and built multi-million dollar businesses. Not a theorist, I am an applied researcher and former Chief Strategy Officer, recognized by Fast Company and Bloomberg Law for my unique work. Through this newsletter, I offer you pragmatic insights into leveraging AI to inform and improve your daily life in legal services.