What happens next for federal workers after AI takes over day-to-day operations?

Bob Benello If you look at AI, not just in the federal government, but in the companies that support it, like the Northrop Grummans and Raytheons of the world, AI is critical to helping them accomplish their missions. Is the mission for the warfighter, or is it for the Department of Veterans Affairs or other areas within the government? And what we’re seeing is a tremendous amount of piloting within these agencies. At the end of the day, AI means different things to different people. And when you really look at it, what are the business outcomes of some kind of AI strategy or AI automation that you’re trying to achieve as a government agency or a federal systems integrator? That’s the key element. When it comes to what AI is, I don’t want to say there’s no magic behind the curtain. AI is happening fast and at scale, coupled with amazing technology from companies like NVIDIA. I consider them the grandfathers of AI. And actually, I think there’s going to be NVIDIA’s GTC event in DC next week. It’s really for governments and you know what AI is doing for governments. So when you look at different areas within different government agencies, automation is key. Having the right large-scale language models to support what these agencies are trying to do, and actually protecting the security of what those AI models are trying to do, and protecting the guardrails that governments should have. This is different from the bad actors surrounding AI testing processes and procedures.

Eric White This is one of the things that interests me about the idea that large scale language models and other AI tools will start to be implemented and contractors will compete for different jobs. How will government officials decide who has a better large-scale language model, or who is piggybacking on whom? Think about what the future of that space looks like.

Bob Benello Well, when you look at this… people say, “Prove you didn’t use an AI model to do that,” right? And I look at that and say, if I’m smart enough to use that AI model, then you should look at me, because I’m innovative and smart enough to achieve my goals. I don’t necessarily think that’s a bad thing. Who is going to leverage the right tools that are out there to get the job done in the most efficient, effective and cost-based space? And I think that’s the key for people to start paying attention. When you actually go for an interview, you’ll find that you’ll be asked whether the interview is being conducted by AI or not. And they have to prove it. But to me, that goes against what AI is trying to do for everyone, right? The key is speed, accuracy and automation. And if someone knows how to make better use of it, that’s probably the person you want because that tool will be built into your environment. That doesn’t necessarily mean it’s a bad opportunity, a bad contractor, or a bad comparison against a large language model. Who is making the most of technology to achieve their goals and business outcomes? That’s the answer.

Eric White How much do you think it will ultimately help you as budgets get tighter and tighter? How much more can this help you?

Bob Benello When you look at what AI can accomplish, a lot of the conversation around AI is around automation. Because automation frees people up from mundane tasks that are simply labor-intensive and allows them to focus on better, more thought-provoking things to do in their environment. Therefore, it makes a difference when it comes to cost. Because if we have an individual that we’re paying $150,000 a year and we’re having him do mundane tasks because that’s part of his job description, now we can automate that and have him do things that are more thought-provoking. It’s good for the environment we put him in, but it’s also good for the bottom line. Because now we can do things faster, more efficiently, and more effectively. And now it can fit within your budget. Therefore, as budgets become increasingly tight, AI becomes a better tool. But, you know, the big worry is… Will AI take our jobs? That’s a very broad question. We shouldn’t let AI take our jobs. If you lost your job to AI, you weren’t really focused on your future career. Because if AI just takes away your job, you’re not building your value as an individual. It’s all about how AI can help us do our jobs better and faster, but what’s at stake now is accuracy, right? Because many mistakes occur within that model. Whether it’s Grok or ChatGPT, all you have to do is ask a question you know the answer to, know the answer they give you is wrong, and say, are you okay with that? And when you prompt them, they say, “Oh, you’re right, actually I made a mistake.” Now that I think about it, this is what happened. So the question isn’t even “Is that model 100% accurate?” If you think it’s the rule of law, you’re going to have a situation where it comes back to bite you in the ass.

Eric White We speak with Bob Venero, President and CEO of Future Tech Enterprise. That’s the selling point of this technology, that whatever the doomsayers say about job loss is wrong, we know it can help automate and free us from mundane tasks. Is that already happening or when will it start? Because I’m still doing a lot of data entry here, Bob.

Bob Benello I think that’s definitely happening. It’s not as efficient and effective as it should be. That’s because we haven’t been properly educated about rapid engineering. If you don’t know how to ask your AI model the right questions, it can take a long time to reach your final results. Therefore, we need to go through a whole education cycle on how to create the right prompts and ask the right questions to reach the end result and goal. And I think that will develop over time. So we are currently in its early stages. I can say that this is definitely useful for some daily tasks. Hey, I want to write about something briefly, this is my topic. That will get you 80% there, but after that you have to go and make adjustments. But that 80% will save you a lot of time and effort from start to finish. But then you need to verify what it is, the final result, and make sure the answer is correct. So we’re not there yet. I think over the next 12 to 18 months, we’re going to see a big difference as these models become more and more intelligent in supporting the businesses and government agencies they serve, whatever their field. Because it’s all about data. And as you know… [garbage] in, [garbage] It’s out, right? This is very important from a data perspective when these models are trained.

Eric White Let’s focus on the defensive side. From a warfighter perspective, can this technology be used in the Department of Defense? Could it also be useful in the procurement contracting world?

Bob Benello Oh, definitely. So when a government agency like the Department of Defense submits an RFQ for some kind of solution, there are often criteria that respondents have to look at and scrutinize every time they do what they’re doing. And those criteria can now be handled by AI, eliminating the need for individuals to compare hundreds or thousands of pages of responses, comb through them, extract important areas from them, and evaluate them against each other. And I think that’s very important. Looking at the speed at which things get done, what we’re seeing with organizations and system integrators is that speed is very important to them. Now it’s important to be able to respond to something accurately and efficiently and be the first to respond to someone else who may not be taking advantage of these tools. If the Department of Defense could use the same tools to compare and contrast with the human eye, things would be very different. That’s really true. It then provides weighted results for each of the potential bidders out there, allowing you to choose what you think is the right solution based on your criteria. But even then, there’s still human intervention to say, “Okay, let’s actually look at the results here.” Thank you for providing the information. The way I look at it, Northrop Grumman has a history of outperforming BAE in this regard, and vice versa, that you can look at from a Department of Defense perspective. So I think the more we hire in this area, the more efficient our agencies will be, the faster we can award awards, and the better our cost base will be. They will also reduce costs.

Eric White Bidders may also be able to use it even if they have to go through a lengthy RFQ.

Bob Benello There is no doubt about it. As a company, Future Tech supports many federal system integrators, so we do a lot of RFQs. And we now have an AI tool – we’ve been in business for 29 years, so we’ve accumulated 29 years of responses – and fed that into a large language model. So when something goes wrong, we no longer have to have a team of seven people say, “Hey, pull from here, pull from there, pull from here, pull from there.” The AI ​​will come in and look at your criteria, take that in and help you draft your responses and key takeaways so far. And this has been amazing both from a time savings perspective and from a people and skills perspective.

Eric White Yes, you gave yourself that example. Now that you don’t have to go through all the documents, what are you making those seven people do?

Bob Benello Here’s a perfect example. Currently leading our RFQ team, she wanted to expand and participate in onboarding and training for individuals joining the company. And now she has a dual role within the organization. That role didn’t exist before, but now we’ve created this additional role. She has both roles and has the time to do it based on the tools that are there. And now we’re going to introduce an AI module to help her in terms of onboarding and training. So if you embrace it properly, it will take you to a good place. I use this analogy all the time. I’m a boater, right? And years ago, when you had two engines, you had two sticks, right? You had left, right, front and back sticks, and now we have joysticks. And the joystick is just as we know it, when you turn the joystick to the left it goes left, when you turn it right it goes backwards and forward. All intelligence is built in. The important and smart thing about this is that I got yelled at a lot. “You’re a fraud, you’re a fraud, you know, you haven’t learned the old ways.” I said, “No, I’m actually smart. It’s much easier this way.” You’ll get to your destination faster, and parking will be much easier. Same thing with these tools, right? Hug them. Implementing these in your environment and leveraging them will help you as an organization. But for the agencies that do it, it’s important right now because it helps them become more efficient and effective, which leads to cost and cost savings.