Artificial Intelligence - I have troubles spelling it!

Well, I asked ChatGPT this question. I think it needs a bit more "learning."

if the ellipsoidal distance between two points is 1000 feet and the combined factor is 1.009, what is the ground distance between the points?

ChatGPT

To find the ground distance between two points given the ellipsoidal distance and the combined factor, we can use the formula:

<math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext>Ground Distance=Ellipsoidal Distance×Combined Factor
</mtext></mrow></semantics></math>

<math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext>
</mtext></mrow></semantics></math>

Given that the ellipsoidal distance is 1000 feet and the combined factor is 1.009, we can plug these values into the formula:

<math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext></mtext></mrow></semantics></math>

<math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext>Ground Distance</mtext><mo>=
</mo></mrow></semantics></math>

<math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mo>1000feet×1.009</mo></mrow></semantics></math>

Ground Distance=1009feet

So, the ground distance between the points is 1009 feet.

Well ya gotta love it. Artificial Nonsense at its finest 🙂

I see absolutely no advantage in introducing AI into our career field unless it involves some mundane task that does not need to be checked by a human with the skill set. We are already automated enough with software routines that take a lot of work off our backs.

Several Attorneys have already been court sanctioned for using ChatGPT to prepare court briefings with the citings being total fabricated and referencing cases/rulings that never existed. The things that we do require brain work because of the complexities of land and water use law and what to apply when under statutes with too many moving parts to be considered by AI.

I see AI as a train wreck in motion.

AI is superb at database applications, so there may be a use in title searches, deed searches and such.

I'm not enough of a visionary to see field or office applications, but there may be some optimization of points or networks application.

Undoubtedly there's some marketing hype that might help sales.

We'll see.

Here is where AI can be used today it has been done way before this latest fad craze. Now this means that those that want to use it they must be willing to put in the parameters themselves not the almighty internet but in house controlled environment That is with extraction of data from point clouds. To be successful you need to understand each band and return signal that your particular scanner has. You need to know the signature of like returns. I think we are very close on the private sector side and manufacturers are already developing it. The one caveat I see is manufacturers are selling globally and a magnolia tree here may return a similar signature somewhere else. The asphalt types the curb concrete mixtures change. But usually those things can be done a different way. I agree with math teacher on the database thing. It can be a tool to use not an end all be all. I think that if AI is implemented over time we will see new terms for GNSS data even in RTK processing in the way receivers handle things. So more software than hardware driven. Kahlman filtering. It is able to learn on its own in a way. Who knows we might end up knowing for almost certainly that if the little box said it’s good then it’s good lol. Maybe.

I think your statement is great. Unfortunately I am doing cut sheets for rough grade curb and gutter storm water line anything that gets a cut fill points all hand entering the data in a spreadsheet. Now my head is about to explode lol. I was cheating a bit until I got my hand slapped and exporting an ascii file and streamlining the data in notepad ++. But thats not the way it’s done. Yes I know I know 8hrs of a dyslexia typing farmer style point number station offset elev design elev description etc etc is more accurate and productive. Sarcasm. Yeah AI is going to change things for sure probably both good and bad. I just hope we learn from history to be cautious and develop the knowledge and skill sets to check it in a sensible manner.

there is absolutely no need to do that in today's world. Most survey related modern platforms allow you to set up a template that you chose the format on and import an ASCII file to do cut sheets with hundreds of points in seconds while also giving you the delta X, Y and Z deltas between the design and staked values for you to look at and see if something was blown in the field.

Oh how I know. Trust me I know. A few years back I was seeing this done for a particular municipality for a water line. They have a specific way. So of course we offset a hub or nail to the water line give cut and fill to the invert. Well this municipality requires an elevation along the profile. I was only working part time for the firm had not used TBC in years and I went down on a rainy day and built a custom report in TBC that did just that. I am no computer wizard either you are so correct no reason in the world to do this. Now you know why my head is about to explode lol. You said something else the deltas. Yes the deltas from the as staked point to the design location. Yeah I am inversing measuring each offset point to its design to check that it’s correct. They don’t allow here cut fills on a stake period . It is a particular area that does this in someway for everything . I have worked with for or been all around the USA and other parts of the world and this was a new thing to me. Heck even when we grades things with a level years ago we had the field book where we kept up with our level run and had our design grades comped so once we got the. New HI we comped rod readings when using a regular level rod not so much when using the linker rod as it was elevation for design grade vs what we read. Which was an awesome tool if used correctly. Loved the linker rod for grading for sure it was fast easy and you could rock and roll.

I agree with your points regarding from a legal standpoint. Would AI get there eventually..Maybe but who knows. I also don't think AI is necessarily the end all be all answer to everything. I think currently it has its place as a general tool that is part of the toolbox. Pick and choose carefully what you use that tool for. Determining what to use the tool for is the part of the issue though. Its like being given a wrench without ever having seen a wrench or understanding it. Some may initially use it as a hammer. But eventually the use becomes evident. AI is the same way. We need to figure out do we use it as a hammer or a wrench. And with AI the 'wrench' will continue to morph and evolve..maybe we can use it as a hammer someday. And a screwdriver and have it make me coffee in the morning.

I think out of the box, AI can be helpful on the office admin/scheduling/project management/HR/data analysis, etc. but not necessarily involved in the actual production field work preparing drawings, doing the actual field work etc.

I continued the conversation this morning. Just another example of never using the excuse "that's what the computer said."

Shouldn't you divide the ellipsoidal distance by the ground distance?

ChatGPT

You're absolutely correct, my apologies for the oversight. To find the ground distance, we need to divide the ellipsoidal distance by the combined factor, not multiply. Here's the correct calculation:

<math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext>Ground Distance</mtext><mo>=</mo><mfrac><mtext>Ellipsoidal Distance</mtext><mtext>Combined F</mtext></mfrac></mrow></semantics></math>

Given that the ellipsoidal distance is 1000 feet and the combined factor is 1.009:

Ground Distance=1000/1.009 feet​

Ground Distance≈990.1feet

So, the ground distance between the points is approximately 990.1 feet. Thank you for catching that!

Editing my reply because I did not your reply correctly.

I thought ChatGPT said your probably correct. haha... but it actually said absolutely not probably. nevermind

I asked CoPilot your exact same question and this is its response:

To calculate the ground distance between two points when you have the ellipsoidal distance and the combined factor, you can use the following formula:

text{Ground Distance} = frac{text{Ellipsoidal Distance}}{text{Combined Factor}}Ground Distance=Combined FactorEllipsoidal Distance​

Given that the ellipsoidal distance is 1000 feet and the combined factor is 1.009, the calculation would be:

text{Ground Distance} = frac{1000 text{ ft}}{1.009}Ground Distance=1.0091000 ft​

Now, let’s compute the ground distance:

text{Ground Distance} = 990.089 ftGround Distance=990.089ft

<cib-overlay>

</cib-overlay>

So, the ground distance between the two points is approximately 990.089 feet.

So... is CoPilot a bit better maybe? interesting because I believe behind the scenes it is the same or based on the same. I have seen this before though where sometimes ChatGPT is better and sometimes CoPilot is better.

One of the characteristics of AI is its ability to "learn" from its experiences. Perhaps ChatGPT and Copilot both learned from the earlier interaction.

Perhaps you or someone else could pose the queston to ChatGPT now and check.

It wouldbe good to have an entity that didn't repeat its mistakes, though.

I actually did post to ChatGPT this morning and it answered correctly as well. Cool... but dang scary too!

Great analogy. So true. This is true with many of the tools we have today in surveying. That is a big part of what we do. We should use the correct tool or best tool for the job at hand. Not always do we have that luxury nor do we always make the wisest decision. I see a lot of today that it is how we always have done it. They will take today’s technology that has many ways to prove it is doing what we need and doing it correctly. Like Chris stated on cut sheets. Yet because we hand entered data many years ago we keep doing it. We take some of the best equipment today yet our process is as if we still log everything in a field book days. I can remember when I moved from only using a field book for logging everything to using the old sdr 24. I ran a total station and I logged every shot in the field book still. Even though we recorded it in the data collector. We did that for several months and go back and down load and check every single point against the field book. Then we knew it was at least recording the information correctly. Then we proceeded to perform all the computations to see if it matched our hand comps. It was a process for sure . When I do something new like discover a new way to get a answer with software. I always go back to check the old fashioned way so I know.

Here's the other side of the question. It seems that ChatGPT can get from A to B after a fashion, but going from B to A ....

If two points on the earth's surface are 1000 feet apart and the combined factor is 1.008, what is the ellipsoidal distance?

ChatGPT

To find the ellipsoidal distance between two points on the Earth's surface given their geodetic coordinates and the combined factor, we need to use the Vincenty formulae or other geodesic distance calculation methods. These formulas take into account the Earth's ellipsoidal shape and provide more accurate results compared to simple Euclidean distance calculations.

The Vincenty formulae are a set of algorithms used for computing geodesic distances between two points on an ellipsoid of revolution (such as the Earth). One of the Vincenty formulas, the inverse formula, can be used to calculate the distance between two points given their latitude and longitude.

However, it seems like you've provided a combined factor of 1.008 without specifying what it exactly refers to. If it's a scale factor, it might be related to a map projection or some local distortion factor, not directly applicable to the geodetic distance calculation.

Could you provide more context or clarify what the combined factor represents?

So true when we have a new more efficient method our gut reaction is to not trust it. We create double the work temporarily to make sure all is good double checking everything. Then, once we are comfortable, we can move toward that better efficiency goal.

Ooops. My bad. I've been away too long.

The combined factor converts between ground and grid, not ground and ellipsoid.

ChatGPT is smarter than I am!!

Yes, instead of combined factor, assume that the elevation factor is 1.008.

ChatGPT

Thank you for the clarification. If we assume the elevation factor is 1.008, it implies that the elevation difference between the two points is being accounted for in the calculation of the distance.

To calculate the ellipsoidal distance between two points on the Earth's surface, including the effect of elevation, we can use the Vincenty formulae or other geodesic distance calculation methods. However, since the elevation factor is provided, we can adjust the horizontal distance accordingly.

Given that the horizontal distance between the two points is 1000 feet, and the elevation factor is 1.008, we can calculate the ellipsoidal distance using the formula:

Ellipsoidal distance=Horizontal distance×Elevation factor

Ellipsoidal distance=1000feet×1.008

Ellipsoidal distance=1008feet

So, the ellipsoidal distance between the two points, accounting for the elevation factor, is 1008 feet.

Poor ChatGPT. I put it in the same situation that an algebra student is in trying to decipher a standardized test question.