Sure. Thanks Craig for the question. Actually, before I answer the question, I think I need to correct a misstatement that I made earlier, although John said it properly, that our cash balance at the end of the second quarter was a little over $40 million. I may have misstated it as $50 million, but $40 million is the correct number, so I just wanted to correct that. With respect to your question, Craig, so more and more of our customers are asking us how we can a help to improve their AI machine learning initiatives and B combine AI and machine learning with our optimization capabilities. The example of the ladder that I often like to use is, AI is good at being predictive. So you could predict demand for products over some future period and then use quantum optimization to optimize the supply chain in support of that demand. So based on these customer requests, we really started thinking through how we could help to our help our customers to better leverage AI and machine learning. And based on work that we had actually been doing ourselves and with some of our customers, what we realized is that we have opportunities in three areas. One is to help our customers build more accurate models by leveraging quantum distributions in the training of those models, rather than classical distributions. Two, develop AI machine learning models with much less energy consumption, because our quantum computers are very energy efficient compared to GPUs, which are used pretty much exclusively to train models today. And then third, as I said, bring together AI machine learning with quantum optimization, and we’ve got some early proof points in each of these areas, as we’ve talked about, some work that we’ve been doing with triumph, which is a leading scientific research institution in Canada, as well as some other companies that are starting to demonstrate the fact that we can be successful in all three of those areas. And then, of course, Zapata AI has some software that actually can be used to get a quick start on doing this quantum AI integration for model training. So essentially, what the roadmap constitutes is a enhancing the lead cloud service with more GPU capability, so that, as soon as possible, our customers can leverage our Leap cloud service, not just for quantum optimization, but also for AI model training, even if only classically to start then be integrating the quantum computations with AI machine learning to kind of enable that integrated machine learning quantum optimization capability, and then finally, actually leveraging the quantum computers in support of doing the model training to be able to deliver better models. So that’s essentially how we are focused on supporting our customers and rolling out capabilities through our Leap cloud service.