The FUTURE of GIS: What to Expect in 2024

2024 is going to be a great year for geospatial just like 2023 was and if you haven't seen my other video and made it here on your own uh you can check that out which wraps up my best things in geospatial in 2023 so go ahead and check that out and then come back to this video or whatever order you want it really doesn't matter anyways so I have four big things that I'm going to watch out for as Trends moving into this next year for geospatial and modern GIS the first one is looking at these comprehensive geospatial or modern geospatial platforms I think we're starting to see some of those really start to come together and be used both in the open source and the clouder proprietary space as well a lot of the cloud providers like I mentioned in my last video are starting to get more complete tool kits things for databases and imagery of course the cloud optimized file storage types are really starting to come together and we're seeing this formation of what you could call the modern geospatial data stack I've talked a little bit about the modern data stack and how that relates to geospatial in a webinar with cardo I'll put the link in the description if you want to check that out it's a full hour long but I think well worth a watch there but uh here's the the kind of like simple architecture I put together together on what the modern geospatial data stack might look like and I think you'll start to see this start to round out as a mix of open tools and maybe proprietary tools and a mix of both of those but really centered around performance and scalability and using interoperability in local settings with things like maybe duct DB and the cloud file storage types and then as you need to you can migrate those parts of your workloads to the cloud that you need to so it gives you this really strong element of choice and things like that but people start to build these really comprehensive deep Cloud platforms that work together and we start to see that even with tools like Leaf map that allow you to connect out these different places the Apache Sedona ecosystem kind of trying to be the data brick sub geospatial with webots and Apache Sedona and the spark ecosystem and then finally things like cardo now I do work for cardo but we've kind of been looking at this space and trying to see it evolve over years but I think you're starting to see these really comprehensive Cloud toolkits really start to take shape so I think we have to watch the space to see how cloud and geospatial really start to evolve over the next year I'll be talking a little bit more about that and some of the ways that you can kind of prepare with different languages or different file types and some of these different things over the course of the next year so definitely check that out now like I said I can't mention sort of this new world without any talking like generative Ai and chat GPT and all these different tools so I think it's a time you might as well talk about that a little bit as well and I think 2024 will be the year that we see some of the first novel geospatial generative AI models now there's really two that I think fall into that category um what I Define as like a novel model it's not just something that you're like running a process and CH GPT or using the GPT or Google apis to run generative Ai workflows and then do something with that but really something that is a Bas net new model that is specific to things like in geospatial that you're training or advancing so the two that I count are the IBM NASA model that actually uses instead of text on the internet imagery that you can actually use to train and use massive amounts of imagery data to train these models to do you know uh land type detection or do different sort of predictions or fire whatever you sort of want from that you can kind of generate and have it scan these new images based off of a ton of imagery that Chang and of course there just massive computing power behind this but uh you can see this demo here that looks just really amazing to see what it can actually do I see that's something to look out for as these like imagery models trained as generative AI models I think that could be really really cool and powerful and see that evolve over the coming year the second is this like sort of goai model I think it's been developed by the University of South Carolina and some others um I'll put all names up here so you can see them so I can cite that correctly but this is one of the first tools I see is sort of yes it uses the GPT apis in the back end uh but what it does is actually creates this like Network effect to saying like okay here's the prompt here's the steps I need to take and then creates multiple other prompts to you know break it down into smaller sets of problems like here's reading in the data here's doing the join here's doing all this instead of having one massive problem then assembles all the steps in the end it kind of looks like something from like a you know a model builder or workflows approach that that you might see and just generated with generative AI so I see these are the first two types and I think this next year we'll see a few more of these sort of Novel generative AI models start to come out and I would keep an eye on this space from there now if you're worried about your job being taken uh I definitely recommend checking out this video from Luke barus who kind of talks about how he thinks that data analyst will actually have more work because of generative AI rather than less work so check that out if you're worried about that I think he has a really good take on all of this now I mentioned the modern geospatial data stack I think that's another thing we'll start to see and this mainly the point here is that if you look at the Modern data stack itself you can see there's a mix of proprietary tools and open tools that you can use to intermingle and using your workflows and I think a lot of people are starting to see that is that you're using open in some spaces you know whether that's open languages or open tools or whatever that might be to actually process and work with your geospatial data I think that's going to translate over into our space and start to have this mix of saying hey I do have some of these proprietary tools but I also want to use these open tools how can we create that sort of mixed system that uh benefits fits the areas that I need I'm getting poor performance here is there something else I could do so I would keep an eye on this as well as this idea of the modern geospatial data stack really starting to evolve and there's a lot of development happening in this areas too I've mentioned the cloud optimized uh file formats for geospatial different things like Duck DB Apache Sedona even using tools like DBT uh lots of different stuff like that there's a lot we can do with this this is one area I plan to explore more over the course of this year is how can these modern data stack tools really be impactful for geospatial work and how can you intermingle those and ultimately how you can learn them and that leads me into my final big prediction for 2024 which is the expansion of continued learning and if you're like me I went through a geography University program I'm sure many of you did that as well or are doing that that's sort of the path that has existed for learning these different tools about outside of a few UD Demi courses or maybe online guides and that's kind of why I started actually doing all of this is because I know there's a really fractured space of how do you learn this different type of stuff I don't there's no road maps there's no clear way to do this different stuff and honestly I think continued learning will start to expand whether that be through courses or certifications or uh more people contributing and writing I'm already seeing that pick up on LinkedIn and sharing these different resources but I think now you'll start to see that as a primary way or a core way of people actually learning these different tools whether that be inside Academia or outside in the workplace so continued learning I know is just kind of growing everywhere it's been booming after the pandemic but I actually start to see this really continuing to grow quite a bit for geospatial in next year so the other cool part about generative AI is actually some of these customized chat GPT gpts I know that's the same thing over and over but I think they're just called custom gpts you do have to have access to the paid version of the open AI chat GPT but there actually are you can program these gpts to do a little bit more than just their C of broad Focus so I've actually put one together here as you can see I've actually put different information into it like a history of my blog post and some of the transcripts from these videos to say here's the way to think about things or here's how to organize these ideas or create a learning plan or stuff like like that it's not perfect but it's out there if you want to try it and give it a shot so um I'll send a link to that and you can actually check that out as well so that's it I think these are the four areas to watch in 2024 uh we can check back this time next year and see if any of these actually came true I have no idea I could uh get you know 100% I could get 0% but I think these are some of the most likely things to happen in this modern GIS geospatial Stace so thanks for tuning in if you want to check out what I thought about 2023 here's the video over here and you can check that out so thanks again for watching if you like this video please like it if you want to subscribe that would be great it always helps with the algorithm so I greatly appreciate it and thank you for checking it out

Channel: Matt Forrest

Views: 6,707

Keywords: gis, data science, spatial data science, spatial data analysis, geospatial, modern gis, matt forrest, spatial analysis, gis analyst, python, python gis, sql tutorial, open source, arcgis, esri, qgis, cloud native, duckdb, aws, google cloud, snowflake, geo news, geo, geo headlines, what is gis

Id: 2WOXQ4JdKaw

Length: 7min 37sec (457 seconds)

Published: Tue Jan 16 2024