I recently came across a research project called ChatGPT which is an artificial intelligence designed specifically to understand written language and to chat online in a very convincing way. So it tries to sound human – great. But it also knows… nearly everything. Because it’s an AI that seems to have consumed most of the internet (although apparently it doesn’t have internet access – presumably so it can’t escape and take over the world).
I asked ChatGPT to write an article about nationwide number portability and IP-based LRNs, using a metaphor to make it easier to understand. This is a topic I started to research a couple of years ago but I haven’t had time to continue my research and share more articles with you. So instead, here’s the article that ChatGPT wrote for me:
Nationwide Number Portability: as explained by ChatGPT
Imagine a world where your phone number is like a bird. Just like a bird, it can fly from one place to another, taking with it all of your important contacts and information. But just like a bird, your phone number can only go so far on its own.
Enter nationwide number portability, the concept that allows your phone number to “fly” from one service provider to another without losing its identity. This means that you can keep the same phone number even if you switch to a different service provider, just like a bird can fly from one tree to another without losing its ability to sing.
But how does this work, you might ask? And how is it different from the number portability that has been possible for the past 25 years?
To understand this, we need to look at how phone numbers are routed. Traditionally, phone numbers have been routed using a system called the local routing number (LRN), which is a 10-digit number that identifies the incumbent local exchange carrier (ILEC) serving a particular geographic area. This system, known as time-division multiplexing (TDM), has worked well for many years, but it has its limitations.
Enter IP-based LRNs, the next generation of phone number routing. Instead of relying on a 10-digit number, IP-based LRNs use a combination of the phone number and the IP address of the service provider to route calls. This allows for more flexibility and reliability in routing, as well as the ability to handle more advanced communication services such as voice over internet protocol (VoIP).
But what makes IP-based LRNs better than traditional LRNs? For one, they provide increased flexibility in routing. Because IP-based LRNs use both the phone number and the IP address of the service provider, they can be used to route calls to different locations depending on the specific needs of the caller. This is not possible with TDM-based LRNs, which are tied to a specific geographic location.
In addition, IP-based LRNs are more reliable than TDM-based LRNs. Because they use IP addresses, they are not subject to the same technical limitations that can cause TDM-based LRNs to fail. This means that calls are less likely to be dropped or disconnected when using IP-based LRNs.
Overall, the transition to IP-based LRNs will bring many benefits for both TDM-based ILECs and their customers. It will enable more flexible and reliable routing of phone calls, as well as the ability to support advanced communication services. And with nationwide number portability, customers will have the freedom to switch to different service providers without losing their phone numbers. It’s a win-win for everyone.
Some thoughts on ChatGPT
This is Andrew again. I’ve been playing with ChatGPT for a few days, and it’s capabilities are quite remarkable. It can write code, you can play games with it, I asked it to diagnose my knee injury, I’ve asked it to help troubleshoot VoIP problems… and I’ve learned a few things along the way.
- It’s really good at answering very specific technical questions about published standards. For example, I asked it to explain the SDP media offer/answer lines on some SIP messages, and it understands the spec better than I do.
- It doesn’t seem to know anything proprietary. Yes I’ve tried asking questions about Metaswitch functionality (I even asked it to write an MMF rule for me), and it either tells me it can’t help, or else it very confidently answers with the wrong information!
- It seems to either be really specific (e.g. technical questions) or else it tends towards being a bit vague. It often uses the high-level information from my question but doesn’t provide more specific details in the answer – just continuing with fairly high-level, vague ideas.
I imagine I’ll continue to have fun playing with this technology, and there seem to be some use cases where it’s incredibly helpful – but (thankfully) it’s not quite so good as to take over all of our jobs just yet. Or maybe I just need to learn how to ask better questions.
If you’ve played with it, let me know what you tried – and be sure to check out these ideas for how to prompt the system.
If you prefer to get support from real human beings, check out our support retainers, as a gift to yourself for 2023!
