****The Official A.I. Thread****

[BigRedChief]

All the new applications and systems made better or replaced with A.I.

First up Google Maps

Introducing Geospatial Analytics

Google Maps Platform is used by over 10 million websites and apps. Customers want to access more insights from our up-to-date, comprehensive geospatial data, so they can make more informed business and sustainability decisions. That’s why for the first time, we’re integrating new Google Maps Platform datasets, along with Earth Engine datasets and capabilities, directly into BigQuery, to make it easier for data analysts and decision makers to access insights. Read this blog to go deeper on our geospatial analytics capabilities. Starting today, the following datasets can be accessed in BigQuery:

Our first Imagery Insights dataset, available in Experimental, applies Vertex AI in Google Cloud directly to Street View imagery in data clean rooms, to identify, detect, and see the condition of key infrastructure objects, like telephone poles, street signs, roads, and bridges. If a telecom company wants to identify which utility poles require maintenance, they can use this dataset to more easily locate all their poles virtually.

Places Insights available in Preview, enables businesses to get customized, aggregated insights about many places in a wider area, based on ratings, store hours, parking, wheelchair accessibility, and more. If a retailer wants to open a new store in an area with a high concentration of expensive restaurants and a low density of large retail brands, Places Insights will provide insights on locations that fit this criteria.

Roads Management Insights available in Preview, helps public sector and road authorities analyze traffic data and improve roads with historical and real-time traffic information. For example, traffic authorities can use the insights from this to identify accident-prone areas and add safety measures like speed bumps or stop signs. They can also build models that forecast traffic conditions to mitigate congestion before it even occurs.

Earth Engine in BigQuery available in Preview, enables organizations to derive sustainability insights, such as wildfire risk or deforestation, from satellite imagery for a specified area of interest. With a new geospatial function and 20 new Earth Engine datasets available in BigQuery, this feature makes advanced geospatial analysis accessible to data analysts–even if they don’t have remote sensing expertise.

In addition to these new geospatial analytics capabilities, Google Earth now provides no-code geospatial analytics tools that enable any professional to access insights. To offer a more comprehensive suite of geospatial tools for developers, data analysts and geospatial practitioners, we're bringing Google Earth–which remains available for everyone–and our new geospatial analytics datasets into the Google Maps Platform family. Visit our new website to learn more about our broader geospatial analytics portfolio.

https://mapsplatform.google.com/reso...i-and-weather/

[|Zach|]

Sora video does some good things but its just....something is missing you have to hold its hand too much. Veo 3 from Gemini looks really good but I have not used it hands on. Veo 2 seemed worse than Sora.

[htismaqe]

Originally Posted by |Zach|

I am sort of settling into a place mindset wise where at this particular juncture. AI won't take your job but someone using AI will.

Yep. That's exactly where we're at right now.

[stevieray]

Like vinyl stickers and drum machines, it has no soul.

[|Zach|]

Data wise it is just fantastic. There is so much stuff you can do but it just knocks out quickly.

I had a hunch some people on my team were getting paid a different rate than expected which is an HR thing.

I had a document listing all of them and their tonsure within our organization. I had another document showing the payscale established by the union showing what someone should be making based on the current fiscal year and their tenure.

Told ChatGPT to compare these two documents and see if these numbers aligned as they should. The comparison showed there were some very small variances that could be attributed to the time of year they were hired but sure enough there was one that was *way* off and it was easy to see that amongst a lot of numbers and people.

This is not something crazy or advanced but it the sort of run of the mill issue that takes less time to sort and is more accurate than a dummy like me pouring over the numbers.

[htismaqe]

Originally Posted by |Zach|

Data wise it is just fantastic. There is so much stuff you can do but it just knocks out quickly.

I had a hunch some people on my team were getting paid a different rate than expected which is an HR thing.

I had a document listing all of them and their tonsure within our organization. I had another document showing the payscale established by the union showing what someone should be making based on the current fiscal year and their tenure.

Told ChatGPT to compare these two documents and see if these numbers aligned as they should. The comparison showed there were some very small variances that could be attributed to the time of year they were hired but sure enough there was one that was *way* off and it was easy to see that amongst a lot of numbers and people.

This is not something crazy or advanced but its a daily thing that was very quick to sort.

Exactly. It can do a "stare and compare" of large data sets in a matter of seconds. Huge time saver.

[Otter]

"create a video of bald men, in kansas city chiefs gear dancing and drinking beer while tailgating before a game"

took under a minute:

https://gifyu.com/image/bsATj

[htismaqe]

Originally Posted by Otter

"create a video of bald men, in kansas city chiefs gear dancing and drinking beer while tailgating before a game"

took under a minute:

https://gifyu.com/image/bsATj

Need to adjust the prompt to make the stadium seats red and yellow and it would be damn near perfect.

[redfan]

I’m seeing a lot of these Super Panavision 70 AI videos of recent movies done in the style of the 1940s and 50s. Entertaining and slightly odd, but some of them look pretty good.
Probably won’t be able to tell the difference in less than a year or so.

[BigRedChief]

Originally Posted by Rain Man

I googled to find out which Hall of Famer was the first to pass away. Google's AI tool told me definitively that it was George Halas. That is dead wrong.

I feel like the AI stuff gives me wrong answers just as often as correct answers. I really have no faith in what AI tells me.

AI is still in its infancy. Think like this is the COBOL/Mainframe era that first launched the tech revolution. Look how far AI has come to practical use of the tech. Mainframe was mostly business. It took another 40 years to get to the internet era. I'd bet money that all the errors and limitations mentioned in here will be gone within a year. Insteaf of 40 years, its being done in 4-5 years.

[BigRedChief]

[

Originally Posted by |Zach|

I am sort of settling into a place mindset wise where at this particular juncture. AI won't take your job but someone using AI will.

AI will initially effect the humans employed in the middle layer of an org. Once its AI running on its own safely, that middle that was gone will start to expand both exponentially upwardly and down the org and your career ladder.

Almost all forecasts say 30% of all jobs in the USA will be gone within 10 years. The IT field that has got so many out of being poor to being able to afford the middle class lifestyle(like me) will be hit harder.

Its going to change the way our current government and society works. Think about it, 30% unemployed. The depression had at its worse, 22%-25% unemployed. The government is going to have to give every citizen some kind of basic assistance to keep people from starving.

[Otter]

Originally Posted by BigRedChief

AI is still in its infancy. Think like this is the COBOL/Mainframe era that first launched the tech revolution. Look how far AI has come to practical use of the tech. Mainframe was mostly business. It took another 40 years to get to the internet era. I'd bet money that all the errors and limitations mentioned in here will be gone within a year. Insteaf of 40 years, its being done in 4-5 years.

It's a shitty time to be a 40 something IT guy. Thank God I'm good cook/bartender and kept life pretty simple and out of debt.

[BigRedChief]

Originally Posted by Otter

It's a shitty time to be a 40 something IT guy. Thank God I'm good cook/bartender and kept life pretty simple and out of debt.

I had planned to work a couple of more years in AI before my memory issues started. That put the kibosh on that idea. Luckily, I was ready enough to retire and not be eating cat food or working restocking shelves in Wal-Mart.

I don't really know you or hitsmaqe but, I'm under the impression both of you are in IT at the upper levels. Some one with experience and good judgment will be needed. There is no way those billion $ businesses are going to allow "Big Balls" to be in control of the companies infrastructure.

[htismaqe]

Originally Posted by BigRedChief

I had planned to work a couple of more years in AI before my memory issues started. That put the kibosh on that idea. Luckily, I was ready enough to retire and not be eating cat food or working restocking shelves in Wal-Mart.

I don't really know you or hitsmaqe but, I'm under the impression both of you are in IT at the upper levels. Some one with experience and good judgment will be needed. There is no way those billion $ businesses are going to allow "Big Balls" to be in control of the companies infrastructure.

Yeah. we already have an entire "responsible AI" organization. There's always going to be people running it, we're never going to just set it loose all on its own. Just gotta make sure you're one of the handlers.

[BigRedChief]

Originally Posted by htismaqe

Yeah. we already have an entire "responsible AI" organization. There's always going to be people running it, we're never going to just set it loose all on its own. Just gotta make sure you're one of the handlers.

Yeah, the "adult in the room" should pay very well.

[htismaqe]

Honestly, Otter probably isn't far off about the English major stuff.

I've found so far that the most effective prompt writers in my org (including me) are the ones that are the most effective written communicators to begin with. The people that send out choppy emails with bad syntax and terrible grammar struggle to write good prompts.