Recently, I was invited to speak on stage at Insight Partner’s ScaleUp:AI conference for a panel aptly titled “Build versus Buy.” It shouldn’t be a surprise that this was a topic of discussion; After all, build versus buy is the great AI debate raging in boardrooms right now—a proverbial fork in the road that may separate the winners and losers of the business world over the next five years.
The “buy” path involves purchasing and implementing off-the-shelf GenAI tools, and if you work in enterprise IT, there are many reasons to love this approach. It’s how you’re already set up to operate. You buy software, integrate it with your workflows and systems, and roll it out to the team. You’ve been down this road before. You know the potholes like the back of your hand.
The “build” path looks a lot more treacherous. It requires building custom GenAI tools for your organization to give you a critical competitive edge. And if you work in IT at most mid-market and enterprise companies, this is not what your team is built to do. Your job isn’t to build new internal products; it’s to ensure that all of the software you’ve bought works well together.
Building and launching new products often requires an entirely different skill set and mindset amongst your workforce. And you’re building with new, unpredictable technology that may hallucinate. That makes this path daunting—filled with unknown dangers that may leap from the bushes.
But in the long run, avoiding the build path may be the most dangerous choice.
Why buying isn’t the answer
Let’s play this out. Let’s say you choose the buy path and arm your team with the latest off-the-shelf GenAI tools—ChatGPT Enterprise, GitHub Copilot, Copilot, Adobe Firefly, etc. And let’s say that you actually train your employees to use these tools. Two things are likely to happen.
First, you likely won’t find that these tools are up to snuff for your use cases. Bain research found that most companies are already building custom applications versus using off-the-shelf tools for every notable AI use case. The reason: They’ve realized that to get the desired value out of GenAI, pure off-the-shelf tools don’t cut it.
This research floored me. Even amongst use cases we think of as solved by off-the-shelf tools—like marketing (ChatGPT, Claude) and coding (GitHub Copilot)—most forward-thinking companies are still building custom, internal GenAI tools to give their teams the efficiency and performance gains they need.
The second thing that will happen is that even if these off-the-shelf tools deliver a productivity boost, you’ll discover two years down the line that you’re stranded without any competitive advantage. It’s the equivalent of having computers connected to the internet in your office in the ‘90s—an advantage for a short period, and then tablestakes.
Without a doubt, off-the-shelf AI tools have value, but if you stop there, you’re not tapping into the wonder of generative AI. You’re not building custom tools tailormade to solve the biggest challenges your customers and employees have, and turn your treasure trove of data into a lasting competitive advantage.
Scale the build barrier
The biggest barrier to GenAI adoption is clear: 62% of corporate leaders say it’s a lack of talent.
After working with dozens of companies to accelerate their GenAI initiatives over the last two years, I’ve noticed a familiar pattern: Many CEOs and CTOs have a misconception about what kind of team you need.
In building high-performing GenAI teams, what we’ve learned is that it’s not about paying a million dollars for a rockstar AI expert who you pry away from the golden handcuffs of Silicon Valley. It’s about bringing on talent and building a team that has a product delivery mindset. People who dissect business priorities and challenges, and then pragmatically solve those challenges with new products and technology.
And while AI deep tech can seem daunting, you also don’t need to build every GenAI component from scratch—many of those building blocks are already available. We’ve seen a 60% boom in open-source GenAI contributions on GitHub in the past year alone. One of the most incredible things I’ve witnessed this past year is how our expert AI community within the A.Team network is able to curate and integrate those building blocks into new AI builds, accelerating the time it takes to get real business value from new AI products.
Inside many orgs that aren’t accustomed to building new products, this is foreign territory, but that’s the beauty of the world we live in today. The workforce has quickly shifted, and many of the most talented AI product builders I know have gone independent and are available to infuse your team with the build-mode mindset you need. After all, the unknown path is a lot less scary when you’re surrounded by the right team.
Raphael Ouzan is the cofounder and CEO of A.Team.
Jelentkezéshez jelentkezzen be
EGYÉB POSTS Ebben a csoportban
The trope of the starving, broke artist has long maintained a place in the public imagination, even as it has morphed into idealized notions of “‘hustle” or “grindset.” “It’s cool to romanticize [
When Apple first introduced MagSafe for the iPhone in 2020, I did not fully appreciate it.
iPhones had supported wireless charging for a few years at that point—and Android phones starte
Even if you’re a regular Alexa user, there’s a good chance you haven’t discovered some of its most efficient features.
Actually, strike that: There’s a good chance you’
The Fast Company Impact Council is a private membership community of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience. Members pay annual
Cryptocurrency exchange Bybit said last week hackers had stolen digital tokens worth around $1.5 billion, in what researchers called the biggest crypto heist of all time.
Bybit CEO Ben Z
Anthropic released on Monday its Claude 3.7 Sonnet model, which it says returns results faster and can show the user the “chain of thought” it follows to reach an answer. This latest model also po
