The AI Boom: Why 2024 is the Year of Enterprise AI Adoption | The Liftoff with Keith

Summary

Episode Summary: "The AI Boom: Why 2024 is the Year of Enterprise AI Adoption" | Liftoff with Keith Newman

Release Date: July 30, 2025

In this compelling episode of Liftoff with Keith Newman, former journalist and Silicon Valley dealmaker Keith Newman sits down with Luke Norris, CEO of Kamuaza, to delve into the transformative surge of Artificial Intelligence (AI) within large enterprises. The conversation unpacks the current landscape of AI deployment, the economic impacts on businesses, the evolving role of consulting firms, and the future milestones set to redefine enterprise AI adoption.

---

1. Setting the Stage: Enterprise AI Adoption in 2024

Keith opens the discussion by highlighting the buzz around AI advancements and their real-world applications. He turns to Luke Norris to gain insights into the actual state of AI deployment within major corporations.

Key Insight: Luke asserts, “This is the year that I think the boardrooms are moving from a test that they could talk to actual execution” (00:40), emphasizing a pivotal shift from experimentation to tangible implementation of AI solutions in enterprise settings.

---

2. Tangible Impact: Cost Savings and Operational Efficiency

A significant portion of the conversation centers on the economic benefits companies are reaping from AI adoption. Luke elaborates on substantial cost savings and operational transformations that large enterprises are experiencing.

Notable Quote: “Our typical customers are seeing cost savings about $20 million within the first month, typically around their ability to, I don't want to say cut, but transform maybe their relationship with consultants...” (01:25).

Discussion Points:

• Reduction in Consultancy Dependence: AI enables companies to transition away from expensive consulting firms, allowing for more in-house efficiency.
• Reinvestment Opportunities: The savings accrued are often reinvested into further AI initiatives, fostering continuous innovation and project expansion.

---

3. The Evolving Role of Consulting Firms

Keith probes into how traditional consulting firms might adapt in response to AI-driven transformations within enterprises. Luke provides a candid perspective on the sustainability of these firms in a rapidly changing landscape.

Notable Quote: “I don't think their business models are truly going to allow them to be the right consultant” (02:09).

Key Points:

• Inevitability of In-House Solutions: As enterprises develop internal AI capabilities, reliance on external consultants diminishes.
• Emergence of New Ecosystems: Partnerships between technology providers and transformation specialists are becoming crucial, as seen with collaborations like IBM providing Microsoft AI services.

---

4. Overwhelming Demand: Scaling AI Implementations

Luke shares an unexpected challenge: the overwhelming demand for AI solutions has outpaced Kamuaza’s capacity to deliver, marking this as the first time in his career that their funnel has become too large.

Notable Quote: “The leads are too much... We've stood up a whole secondary company practically just to handle the influx” (03:53).

Discussion Points:

• Forced Slowdown: Despite the high demand, operational constraints necessitate a slowdown, prompting the search for reliable partners.
• Need for Strategic Partnerships: Identifying and collaborating with technology transformation partners is essential to manage and scale AI implementations effectively.

---

5. The Ideal Partners: Bridging the Gap

Delving deeper into partnership dynamics, Luke defines the characteristics of ideal partners to support AI deployments.

Notable Quote: “We're looking for technology transformation partners. So somewhere between the typical value-added reseller and the Deloitte” (04:26).

Key Characteristics:

• Middle-Tier Firms: Companies that operate between hardware resellers and large consulting firms, such as WWT government acquisitions.
• Capability in Delivery: Partners must possess robust delivery mechanisms to handle extensive AI projects within enterprise environments.

---

6. Beyond Training: The Rise of Inferencing in AI

The conversation shifts to the technical evolution within AI, highlighting a significant trend towards inferencing—the application of trained models to make decisions or predictions.

Notable Quote: “2000, 5000 enterprises are going to adopt inferencing, actually running AI now that's going to be the meaty center” (05:01).

Discussion Points:

• Shift from Training to Deployment: While training AI models garners media attention, the real value lies in deploying these models at scale within businesses.
• Role of VARs: Value-Added Resellers (VARs) are pivoting to focus on the deployment and maintenance of AI inferencing, filling a critical niche in the AI ecosystem.

---

7. Future Horizons: Emerging AI Models and Tools

Concluding the episode, Luke provides a forward-looking perspective on upcoming AI advancements poised to revolutionize enterprise applications.

Notable Quotes:

• “Qin 3.0... exceeding GPT-4, Gemini, even the models for Microsoft and OpenAI” (05:42).
• “The MCP model context protocol... agents can access anything with the best models in private ecosystems” (05:42).

Key Developments:

• Open-Source Dominance: Models like Qin 3.0 are setting new benchmarks, challenging established giants like OpenAI with superior performance.
• MCP Model Context Protocol: This innovation enables AI agents to interact with an expanding array of tools seamlessly, forecasting a near-exponential growth in accessible AI functionalities.

---

8. Closing Thoughts: Navigating the AI Boom

As the episode wraps up, Keith and Luke reflect on the exhilarating yet challenging journey ahead for enterprises embracing AI.

Notable Quote: “What a good problem to have” (06:57).

Final Insights:

• Optimistic Outlook: Despite operational challenges, the overwhelming demand for AI solutions underscores the technology’s transformative potential.
• Sustained Growth: With the right partnerships and continuous innovation, enterprises are well-positioned to harness AI’s full capabilities, driving unprecedented efficiency and innovation.

---

Conclusion

This episode of Liftoff with Keith Newman offers a deep dive into the seismic shifts underway in enterprise AI adoption. Through Luke Norris’s expert insights, listeners gain a comprehensive understanding of the economic benefits, the changing consulting landscape, the importance of strategic partnerships, and the future trajectories of AI technology. As 2024 unfolds as a landmark year for AI in large organizations, the conversation underscores both the opportunities and the logistical challenges that come with this transformative wave.

---

Timestamp Reference:

• 00:40 - Shift from AI testing to execution in boardrooms
• 01:25 - Cost savings and transformation with AI
• 02:09 - Sustainability of consulting firms
• 03:53 - Kamuaza’s overwhelming demand
• 04:26 - Characteristics of ideal AI transformation partners
• 05:01 - Rise of inferencing in AI applications
• 05:42 - Emerging AI models and the MCP protocol
• 06:57 - Positive outlook on high demand challenges

---

Listen to the Full Episode: Find this and over 80 other episodes of Liftoff with Keith Newman on Apple Podcasts.

Transcript

[A]

All right, we have Luke Norris, the CEO of Kamuaza. Yeah.

[B]

You nailed it.

[A]

Well, I never played the video game, as you told me. I'm a little jealous and want to check it out now.

[B]

It is. It's pretty good, to be honest. Yeah. Wow.

[A]

Outside of Breckenridge, Colorado, you're calling home. That's fantastic too.

[B]

I am blessed beyond words. Yes.

[A]

Well, that's great, Luke. Hey, give me a level set out there. I love talking to CEOs and founders about where we are today in the AI revolution, if you will, in terms of deployments. You hear a lot of people talking about things, but where are we in terms of real world deployment?

[B]

So this is the year, and I'm not saying that to be optimistic. I'm saying that because we have over 20 Fortune 500s and global 2000s running significant production workloads. This is the year that I think the boardrooms are moving from a test that they could talk to actual execution. Also, the macroeconomics that are happening right now, they're trying to find ways to buffer themselves and to sort of get more efficient at less cost. I just did the keynote for intel at Intel Vision, where we actually did a live demo with Department of Homeland Security, a massive scale project on them doing emergency preparedness all via generative AI. So I really believe the software and the hardware are finally at the point the enterprise can do inferencing the at extreme production workloads and really get the value out of it.

[A]

So that means we're seeing case studies, stories about significant cost savings now that are showing up in the bottom line.

[B]

So our typical customers are seeing cost savings about $20 million within the first month, typically around their ability to, I don't want to say cut, but transform maybe their relationship with consultants so they no longer need the Deloitte's and the Andersons that are doing very costly transition level work, our modeling work, AI now can take that over overnight. And once you get that level of cost savings, you have the money to reinvest into it and really start to get some cool projects going.

[A]

Right. So would we see the consulting firms then just to kind of play it out up, leveling their skills and happily utilizing these services directly or indirectly?

[B]

You know, probably for a drink after this, for you and I to talk through on that. I don't think their business models are truly going to allow them to be the right consultant.

[A]

But what I think about it was that I just read where IBM's going to be providing Microsoft services around AI solutions. It seems like we're creating some new ecosystem partnerships.

[B]

Yeah, I mean, they have to. The enterprise is asking not just for the efficiency gains of AI, but the cost gains of it. So you can't just say, hey, Deloitte or whoever, help me on the efficiency side. You also have to have that transference of the cost side. And that really means a lot of it has to eventually be done in house. But they need people to help them transform, and that's where we're seeing these new ecosystems. They're just helping companies fish. And once they learn to fish, they're off and running on their own.

[A]

Right. Teach a man to fish.

[B]

You got it.

[A]

The next point I would ask is. So you're sort of winning over the early adopters.

[B]

Actually, no, let me stop you right there. The Fortune 500 and global organizations are adopting this for the first time as the first technology that they're the early adopters. So one.

[A]

Well, I don't define them as.

[B]

I just want to be clear. One of our customers, Campbell Soup, 150-year-old, 8 year old company, I think they're the second or third oldest on the Fortune 500 list. They're in a mass adoption of generative AI right now. So it's these laggards that would typically be seen as laggers are actually the ones transforming.

[A]

What I was referring, Luke, to you. It captured like 20% of the 500.

[B]

No, not quite.

[A]

So I was saying.

[B]

I said about 20.

[A]

Okay, well, I use different math, so you captured that part of it. Those are the early adopters. Now how do you get the rest? How do you continue to grow your funnel and help accelerate these pipelines on the next year?

[B]

So first, it's my fifth startup. It's the first time in my entire career the funnel's too large.

[A]

Yeah.

[B]

The leads are too much.

[A]

I hate when that happens.

[B]

Isn't it amazing? We are actually in a forced slowdown because of implementations with our customers. And we're actually trying to find more and more partners that can help us with those implementations so we can actually move on with adopting and actually helping more organizations, especially in the Fed space. Literally, the funnel is so large, we almost can't handle it. We've stood up a whole secondary company practically just to handle the influx.

[A]

That's really interesting to me. So what do those partners look like? I guess it's not Deloitte.

[B]

You know, we're looking for technology transformation partners. So somewhere between the typical value added reseller that would sell the hardware and the Deloitte. So there's A really nice middle firm and the MSP and the MSP type scenario. So like WWT government acquisitions, these are amazing companies that fit in between the box set and the service provider and they're doing amazing delivery.

[A]

There's a little vertical of government reseller integrator types. Right. That serve that market. Are you seeing that ecosystem changing to adopt to the AI demand?

[B]

Yeah. So AI is fueled so much right now on the training side and there's just billion dollar after $10 billion sales and they make the big headlines. But it's the fact that 2000, 5000 enterprises are going to adopt inferencing, actually running AI now that's going to be the meaty center. And we're finding the vars that maybe couldn't contribute in those billion dollar sales are now focusing on that inference, that middle layer and they're transforming their own organizations to sort of take advantage of that.

[A]

One last question and this has been great conversation but you know, you're a busy guy. So what are we going to see as the next big step, the next big milestone in terms of the models that are coming to market, the tools that are coming to market?

[B]

Yeah. So two things on that. On Monday, Qin, the Alibaba research group released Qin 3.0 private open. Sorry. It's a public open weight model. So anybody can use it. All of our enterprises can use it. It is now the number one model on all test bench across all levels. It is out exceeding GPT04, it is exceeding Gemini, it is exceeding even the models for Microsoft and OpenAI etc. So it's absolutely amazing that now open source models have taken that leap and they're at the soda level. They're the state of the art, just like the other ones. Second, the big thing, and I'm sure you've had 10 people already say this MCP model context protocol, the ability to write once that an agent now can access this tool. This tool can be anything. You can write that now you have a protocol for that to actually happen. And once one company writes that tool, they open source that MCP agents now can access. I think the last I saw on the search engines was about 150 to 200,000 tools. I bet you just give it another three months, they'll be in the millions. So now agents can access anything with the best models in private ecosystems, that is their public models. You can do this. It's going to change the enterprise and change the world.

[A]

That's exciting times. It is. Well, I think you're going to have a hard time delivering all this demand that you have coming your way.

[B]

What a good problem to have.

[A]

Finally. I was going to say the pity in my heart is not that heavy.

[B]

I appreciate that.

[A]

Thanks so much for your time by.

[B]

All right, thank you.

[A]

Have a great show.