103 | 👀 📰 The Other AI M&A Deal

In today’s Brainyacts we:

1. highlight a top referrer of Brainyacts

2. cover a big day in M&A

3. provoke your thinking on contracts

4. talk ‘learnright’ vs copyright

5. get optimistic about AI with a VC

6. learn US House of Representatives has to use paid version of OpenAI

7. see OpenAI coming for Microsoft’s Co-pilot

8. get classroom/teacher-specific AI models

9. glimpse the near-term Star Trek universal translator

10. cringe over the crappy t-shirt Biden gave India’s PM

👋 to new subscribers!

To read previous posts, click here.

🌋 🔗 Big day for AI M&A

A Pragmatic Perspective on AI in Legal Markets: Spotlight on the MosaicML Acquisition and Its Contrast to the Casetext Deal

Last night’s and today’s headlines about Thomson Reuters acquiring legal tech startup Casetext have rightly dominated discussions on generative AI's impact on the global legal market today. While this high-profile acquisition undoubtedly signals a transformative shift, another noteworthy development—the Databricks-MosaicML merger—shouldn't be overlooked. Announced yesterday as well, the acquisition has profound implications for the legal industry, particularly concerning private custom AI models, a feature increasingly sought after in legal tech.

Thomson Reuters’ acquisition of Casetext for $650M is undoubtedly a game-changer for the legal industry. Casetext’s AI assistant CoCounsel, developed in partnership with OpenAI, is designed to streamline legal research, document review, and contract analysis. By incorporating this AI-driven tool into its portfolio, Thomson Reuters is set to reshape how legal professionals work, promoting efficiency, and potentially democratizing access to quality legal research.

However, the model of this AI tool is fundamentally universal, intended to cater to all legal players, and does not allow for specialized, proprietary AI tools customized for specific needs or client bases.

👉 On the other hand, Databricks' acquisition of MosaicML for $1.3B signals a shift toward empowering entities to build and run their own custom machine-learning models.

In stark contrast to the Thomson Reuters-Casetext deal, the Databricks-MosaicML merger addresses a burgeoning demand within the legal industry (and beyond): the need for privacy and customization in AI tools. As legal firms deal with sensitive information, a one-size-fits-all AI solution may not suffice. Databricks and MosaicML are offering a solution that could potentially address these concerns.

MosaicML offers cost-effective AI tools that enable businesses to have full control over their data used to train AI models, thereby mitigating data privacy concerns and ensuring proprietary information doesn't leak to competitors. Its acquisition by Databricks signifies a step toward providing law firms the ability to create specialized AI models tailored to their unique needs and clients. This could revolutionize how law firms handle internal data, client communications, case analysis, and more.

The potential implications of these deals underscore the importance of discerning the unique value proposition of different AI technologies.

While Thomson Reuters-Casetext presents a novel, comprehensive tool that can enhance legal research across the industry, Databricks-MosaicML taps into the niche demand for custom, private AI models, with a promise of autonomy, privacy, and specificity.

As we navigate the intersection of AI and law, it's imperative that the legal industry casts its net wider and explores developments beyond the confines of legal tech.

Indeed, while the Thomson Reuters-Casetext deal presents significant advancements in legal AI tools, the Databricks-MosaicML merger serves as a compelling example of beneficial solutions arising from the broader generative AI market. Legal professionals must recognize that not all necessary solutions will be built exclusively within the legal sector.

By monitoring the wider AI ecosystem, they can identify and harness versatile tools, such as those provided by Databricks and MosaicML, to create custom, private AI models that cater to their unique needs. The future of legal practice calls for openness, adaptability, and a broad perspective, embracing the comprehensive AI landscape to realize the full potential of this transformative technology.

Resource: The deck that TR released upon its Casetext announcement

📑 🔮 The Future of Contract Lifecycle Management: A Generative AI Perspective

I am lucky to get connected to so many cool and interesting people through this newsletter. Yesterday I had a fascinating conversation with a head of legal and compliance who has been in-house counsel for a few notable companies. We were discussed the opportunities in the contract lifecycle management space - get all geeky and riffing on the potential for generative AI to improve the holistic experience of contracting, not merely the process of contracting.

It got me thinking of some problem areas/unmet needs that many overlook but that generative might be able to address.

📧 Would love your feedback and reactions! Just hit reply to this email to send me a note.

The conventional lens of Contract Lifecycle Management (CLM) has been mainly focused on the stages of contract formation and drafting negotiations. However, this viewpoint overlooks the multifaceted cognitive aspects underlying these activities. By paying attention to these often-neglected areas and utilizing the transformative potential of Generative AI, we can usher in a new era for CLM that doesn't merely enhance contract performance but also humanizes the process.

AI-Assisted Assessment of Bargaining Power: In the current CLM landscape, the assessment of bargaining power is largely dependent on human intuition and experience, a process that is prone to bias and error. Generative AI can revolutionize this aspect by harnessing real-time data from communications, news updates, and prior interactions to continually assess the bargaining power of all parties involved.

For example, a company about to negotiate a major contract with a supplier could use Generative AI to discover that the supplier has recently lost some key clients and is keen on securing new business. With this insight, the company could negotiate for more favorable terms, armed with the newfound understanding of their increased bargaining power.

Automating Communication Cascades: Contracting is a collaborative process involving a torrent of emails and communications to clarify business points, update stakeholders, and decode contractual details. Generative AI can streamline this by drafting, organizing, and automating these communications.

Picture a scenario where several emails must be sent to different stakeholders during the finalization of a contract. The Generative AI could automatically draft these emails, tailoring the language and content based on the recipient's role, saving the legal team's time and ensuring consistent, clear communication.

Simplification and Translation: Legal jargon, while necessary, can create a disconnect between the legal team and other stakeholders. Generative AI can bridge this gap by translating complex legal terms into plain language that everyone can understand.

Consider a new contract that the project team must comprehend to fulfill its obligations. The Generative AI could generate a summary of the contract in plain language, ensuring all team members understand specifics such as delivery timelines, penalties, or performance metrics.

Proactive Risk Mitigation: Generative AI could provide a proactive approach to risk mitigation. By analyzing historical data, industry trends, and clauses from thousands of similar contracts, it could suggest modifications that minimize potential disputes or liabilities.

In the context of an international deal, for example, the Generative AI could flag the lack of a common clause to manage foreign exchange risk in the contract draft. This foresight could help the company mitigate potential future risks and address them before they escalate into problems.

Continuous Improvement and Learning: Generative AI could play a significant role in post-execution contract management. It could flag non-compliance, suggest areas of improvement for future contracts based on historical data and performance metrics, and even automate reminders for crucial deadlines or obligations.

Let's say a company has a history of facing penalties due to non-compliance with certain contractual obligations. The AI system could identify this pattern, recommend preventative measures for future contracts, and foster continuous improvement.

Contracting Forecasting: The learning capabilities of Generative AI could be harnessed for forecasting future contract needs based on past trends and business growth.

For instance, in-house counsel planning for the next quarter could use Generative AI to predict an increase in contract needs based on historical data, allowing the company to allocate resources, hire additional support if needed, and better manage their workflow in advance.

As the legal landscape evolves, in-house teams are uniquely positioned to pioneer this generative AI revolution in CLM. More than just a technology upgrade, this transition signifies a shift in perspective, a recognition of the oft-unseen human factors influencing the contracting process. It is a call to be more cognizant of the needs of our teams, our internal clients, and our contracting partners.

It is not enough to stay on the well-trodden path of traditional CLM. We need to embrace the transformative power of Generative AI and, in doing so, redefine the contract lifecycle. We must build a future where contracts are not just legally sound, but also strategically effective, intelligently crafted, and human-centered.

🤔 📜 AI, Copyright, and the ‘Learnright’ laws?

As AI continues to learn, create, and even generate its own content, we're entering uncharted waters in the realm of copyright law. What happens when machines can replicate the style of our favorite artist, or paraphrase our most trusted news source? Are our current protections enough?

MIT's Thomas Malone introduces the concept of "learnright", a potential solution for balancing the advantages of generative AI with the rights of content creators. But like any sea change, it's not without its challenges and critiques.

How should learnright protection be invoked? What implications could it have on the progress and diversity of AI development? Are we at the dawn of a new era where legal frameworks and AI technology can coexist harmoniously?

🎯🏆 A law firm with the right POV on Generative AI

Marc Andreessen, the trailblazing technology entrepreneur best known for founding Netscape and co-founding the esteemed venture capital firm, Andreessen Horowitz, has recently taken to the written word. In a thought-provoking article, Andreessen explores and provides his unique insights into the evolving world of technology and business.

Andreessen later appeared as a guest on the Lex Fridman podcast, where he had the opportunity to delve deeper into his point of view and expand upon the concepts introduced in his article. If you have the bandwidth, I highly recommend both reading Andreessen's piece and tuning in to his episode on Fridman's podcast.

For those in a hurry, I'd like to spotlight a particularly noteworthy segment from the podcast - a concise 60-second clip that is absolutely worth your time. In this excerpt, Andreessen shares his initial assumptions about applying generative AI within the legal sphere. Like many, including a good number of legal professionals, he surmised that this technology would primarily serve to validate facts, cases, and laws.

However, Andreessen found that a particular law firm (whose identity I'd be very curious to know) offered a refreshing perspective that deviates from this expectation. They see generative AI not just as a fact-checking tool, but as a creative instrument - a view that, in my humble opinion, hits the nail right on the head. Without further ado, here is that intriguing clip.

In the Memetime: 

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