Generative AI: FOMO or FUD?

In the work that I do, I have the privilege to speak with many companies about a wide array of issues. Increasingly that includes their AI strategy. And one thing that I have been amazed to learn is how many are building in-house solutions for AI.

The focus on AI is not surprising. OpenAI, probably the most recognized name in Artificial Intelligence right now, has been around since December 2015. Since then, the global AI market has mushroomed from an estimated value of $12.75 billion to nearly $94 bn in 2022 (Statista, July 2023).

Generative AI is following a similar trend. According to CBInsights, in 2019, equity funding in generative AI startups sat at $2.9bn. By 2023 that figure had increased to nearly $22bn. And according to a recent McKinsey survey, 79% of companies surveyed have had some exposure to generative AI, and 22% are regularly using it in their work.

So, to hear that more companies are baking AI into their operations and strategy is in line with overall trends. What surprises me is that so many companies are building their own solutions – even ones who are not in the business of building technology.

FOMO OR FUD?

What is driving the investment of time and resources into this in-house development? Is it FOMO (Fear of Missing Out) or is it FUD (Fear, Uncertainty, and Doubt) about the existing products in the market? My guess is that it is both. On the one hand, FOMO drives a desire to participate in technological advances and there is a real fear that by not doing so, it will impact their business negatively in some way. While FUD argues that there is too much security risk to do so using existing third party platforms.

In the case of AI, lawyers, regulatory affairs, privacy, information security, and compliance teams are right to raise flags and identify the risks involved with adopting new technologies. Generative AI can be a great tool, but it is still developing and early uses almost certainly involved the sharing/exposing of confidential information. In addition, hallucinations - anomalies where AI systems produce unexpected or incorrect outputs - have resulted in fake legal citations, and great sounding, but ridiculously wrong accounts of factual circumstances. Those professionals focused on minimizing risk are doing their jobs well when they evaluate and sometimes refuse to green light the adoption of “cool”, new technologies.

Therefore, the option of building an in-house solution can seem like a good idea. But is it?

FOMO BECAUSE OF FUD

It’s the same thing I saw in the days that public cloud technology was the “Generative AI” of the day. Everyone was excited about the revolutionary nature of this new way to generate computing resources. No one wanted to miss out on the opportunity. But in many corporations, regulatory, legal, compliance, and infosec teams identified significant risks. Regardless of whether such risks were accurate or not, roadblocks were thrown up. FOMO argued for the opportunity to move forward. But it was FUD which led all sorts of companies to build their own solutions. I recall working on deals to create in-house clouds with players ranging from pet supply companies to the biggest names in banking.

Years later, we see very few companies truly operating their own cloud infrastructure. That’s largely because technology has advanced and players like AWS, Google Cloud, and Microsoft Azure provide solutions that allow the regulatory, compliance, legal, and infosec folks to sleep at night. But it’s also because in-house solutions themselves can present even greater business risk.

CMF REQUIRED

Building a technology internally requires CMF - Care, Maintenance, and Feeding. Any internal technology system needs to be monitored, maintained, and have a proper support network in place for it to work long-term. If you are not in the business of the care, maintenance, and feeding of technology solutions, your in- house AI solution is probably a poor investment.

I have asked some companies about their CMF plans, should their solution be an effective one. Largely the answer that I got was, “we’ll figure that out when we get there.” But here’s the thing, these are companies with incredibly smart people. I would say many are much smarter than I. So why are they doing this?

An in-house solution that is popular and funded today, may become lacklustre down the road. If the team engaged in that project are no longer able to continue to support it, then the solution may not only become buggy or outdated, but it also can become a security, compliance, and business operational risk. Without a CMF plan, that great in-house solution could turn into a major liability.

SO, WHAT’S THE ANSWER?

If your company isn’t signed on to CMF well into the future, is the only option to actually miss out?

No. There are several companies out there with great solutions for using Generative AI in a manner that doesn’t require a breach of confidentiality or exposure to outlandish internet-based hallucinations. Companies like Accubits and Solulab. Even better, there’s no CMF required. These companies will not only undertake the CMF for you, but they provide specific solutions that take into account customers’ legal, compliance, information security, and regulatory obligations.

Accubits, for example, uses generative AI to automate a broad range of corporate services, whilst maintaining a core commitment to what they term “Responsible AI”. In the RFP space, RocketDocs launched the strategic response industry’s first private AI solution designed to enable use of generative AI while maintaining confidentiality and an effective approvals process for newly generated question/answer matches.

No matter what industry you are in, you needn’t worry about missing out. Whether it is aviation, aeronautics, life sciences, pharmaceutical, financial services, or healthcare, if advancing your company while maintaining compliance obligations is your goal, there’s a solution out there. The important thing is to work with companies that work for you. If you’re not sure who provides the solution that you need, reach out to me. I’ll look for a company that meets my high standards and I’ll introduce you.

The end result is getting to adopt new technologies while minimizing risk.

To be clear, this concern is not unfounded, especially in industries where data privacy and regulatory compliance are non-negotiable. Some companies even conducted thorough due diligence on potential suppliers offering generative AI solutions, only to discover that most leaders in the industry failed to meet their requirements. This often left these companies with the impression that they had to develop their own solutions internally.

In light of this, companies are faced with a dilemma: either risk non-compliance with their regulatory and legal obligations by using third-party AI solutions or take on the daunting task of developing in-house systems that might not align with their core business competencies. The irony here is palpable. By building an in-house solution, companies are inadvertently increasing their risk profile, especially in terms of non-compliance, simply because maintaining and evolving an AI system requires resources, expertise, and a level of technological agility that many organizations may lack.

The decision to build in-house often stems from a fundamental misunderstanding of the complexities involved in developing and maintaining AI technologies. It’s not just about coding and deploying an AI model; it’s about ongoing maintenance, continuous learning, ensuring data integrity, and adhering to ever-evolving regulatory standards. For many companies, this means diverting attention and resources (i.e. money & people) away from their primary business operations, which can lead to unintended consequences, including reduced competitiveness in their core markets.

FOMO With AI

Today, I want to pivot the discussion to another critical aspect of the AI industry: the current frenzy of investment in new AI companies. While it’s true that there are many promising new AI companies with innovative products, I have observed a troubling trend. A significant number of companies with impressive Series Seed and Series A fundraising rounds are entering the market with no novel technology and, in many cases, incomplete solutions. These companies are securing millions in funding, not because they have groundbreaking ideas, but because of the pervasive FOMO—Fear of Missing Out—among investors.

Venture capital firms, driven by the fear of not having an AI play in their portfolios, are making hasty investment decisions. In some instances, I’ve seen companies raise $15, $20, even $60 million simply by taking an open-source AI technology and repurposing it for an existing industry. The catch? Their solution often addresses only a fraction—sometimes as little as 10%—of what the industry’s customers actually need. This trend is not only a potential misallocation of capital but also a disservice to the companies that will ultimately rely on these underdeveloped solutions.

The implications of this investment behavior are far-reaching. For one, it inflates the perceived value of AI technologies that may not be ready for prime time, leading to a bubble that could eventually burst, leaving investors and customers in a difficult position. Moreover, the hype surrounding these newly funded companies often overshadows more established players who have been steadily innovating without fanfare. These startups, flush with cash, often engage in aggressive marketing campaigns, making bold claims about the capabilities of their products—products that, in reality, may still be in their infancy.

Our Role In What’s Coming

RocketDocs celebrated its 30-year anniversary, and while we may not be the hottest new name in the AI startup scene, we have something far more valuable: decades of experience and a deep understanding of the industries we serve. We have not been chasing down venture capitalists for investments, nor have we felt the need to ride the wave of AI hype. Instead, we’ve focused on building robust, reliable, and innovative solutions that genuinely meet the needs of our customers.

Our approach is guided by a core principle we call Responsible Release. This philosophy, which someone much smarter than me coined, is rooted in the belief that we should not release a new product until it is fully ready for customers to use. This might sound like common sense, but in an industry where speed to market is often prioritized over product readiness, it’s a principle that sets us apart. Our business is focused on customers who need technology to assist them in answering RFPs, RFIs, DDQs, InfoSec questionnaires, and more—with accuracy and reliability.

These customers, who operate in regulated and highly competitive industries, require confidentiality and precision. They need solutions that match questions with correct, approved answers and a technology provider that doesn’t send their data into the wild, potentially generating responses that look good but are fundamentally inaccurate.

At RocketDocs, we know the strategic response and RFP response industry inside and out. Our team members bring decades of experience, having sat in the same seats as the customers they now serve. This deep industry knowledge is what informs every product we develop and every decision we make. When we introduced our generative AI solution, we did so with our customers’ specific needs in mind.

We launched a private AI solution that takes a hybrid approach, pairing our existing AI matching technology with a private generative AI model. This hybrid approach ensures that the solution not only meets the highest standards of accuracy and confidentiality but also integrates seamlessly into our customers’ existing workflows.

A key feature of our solution is its approval workflow. We understand that subject matter experts, or SMEs, play a crucial role in maintaining the accuracy and integrity of content, which is why our solution makes it easier for SMEs to approve newly generated content before it is added to their content libraries. This ensures that our customers can trust the outputs generated by our AI, knowing that every response has been vetted and approved by the right experts.

Substance Over Style: The RocketDocs Way

Are we a shiny new startup with a flashy marketing campaign and a pile of venture capital? No. We are an established company with a focus on substance over style. We do a lot with the resources we have, and we take pride in building great products that truly serve our customers’ needs. We may not be in the business of making flashy headlines, but we are in the business of delivering reliable, effective, and responsible technology solutions.

If you’re watching the AI space, you won’t see venture capitalists (VCs) shifting their focus from startups to 30-year-old companies. VCs have their way of doing business, which often prioritizes quick wins and high-growth potential over long-term stability and responsibility. But at RocketDocs, we have our way of doing business too—one that prioritizes the needs of our customers and the integrity of our products. And that is what we will continue to do: provide the best strategic response technology in the industry, with a focus on delivering real value, not just hype.

The burgeoning AI “industry” is at a crossroads. On one hand, the excitement and potential of generative AI are undeniable. On the other hand, the rush to the market and the influx of investment are creating a landscape where FOMO often outweighs thoughtful, responsible innovation. At RocketDocs, we believe that the true value of AI lies not in being the first to market but in being the best at serving our customers’ needs. As the industry continues to evolve, we will remain committed to this principle, ensuring that our solutions are not only innovative but also reliable, accurate, and, above all, responsible.

- Perry Robinson, RocketDocs CEO