We're in Trouble - How the Use of AI Search may Hasten Our Societal Demise

 I was just reading an article about Google and ChatGPT.  If I understand the article correctly, in times past, Google was hesitant to release AI driven products, for safety reasons.  Now, due to the threat of products like ChatGPT and other AI creation tools, Google is softening its stance on its tools and preparing them for release (Ref link). Here's where the problems come in.  Depending upon the social/economic class that you are apart of here in the US you may see things differently.  You see as I've had more experience and learned more throughout this life I realize that we (society as a whole) have put our trust and dependencies into a bunch of half baked products.  

Let me give you an example.  Just yesterday I was making my first reel post on Instagram and wanted to sync/edit some photos (three to be exact) to go with a clip of music.  I have been editing video since 2004 so I know my way around.  Anyway, so I go to edit these photos with the music on IG and it guides me to use the Auto Sync function.  So I'm like cool, the app will automatically sync the music and photos up for me, great!  I hit okay and the process runs.  Turns out that I don't like the duration that the program used for the first photo.  Easy enough to fix, so I thought.  I tried in vain to intuitively use the trim controls to lengthen and the first photo, but the trim/edit tools seemed archaic.  I won't bore you with all the details, but let's just say it caused me to yell out loud several times during my attempt.  I said to myself, "clearly the person/people who created this have never used it".  I know that this statement is (or should be) incorrect.  Of course they used the app, you have to when developing software.  What is probably true is that they've never used the app as an individual with video editing experience who's in a hurry and has to quickly post a reel.   

What I'm alluding to here is perspective.  I believe that many of the people who are hired by these companies as software developers do just that, develop software.  They are not necessarily end users.  They will often times write the code to develop features that are efficient for them, but not for the end users actual use.  The Managers, who aren't end users either, approve these features because all they are really concerned with are deadlines and budgets.  So if this happens over and over again on mundane low risk applications, what do you think is going on with the programmers developing Chat AI's?  

Here's another example about perspective.  A long while back I wanted to create a fake profile (the results are in the image atop this article).  So I was aware of some online sites that have AI generated profile pictures.  Me being an African American/Black male, I wanted an AI generated picture to reflect me (or someone like me).  So the site would allow the user to specify all the characteristics and then create the photo.  How about the photos that came  out for a dark skinned short hair man all looked like he was from India (no disrespect to all my Indian colleagues).   Myself and my profile avatar are African American/Black and there's a difference.  I contacted the site about this issue, they acknowledge it and mentioned that they are trying to increase their data sample size (update 1/23, nothing has changed).  

Back to ChatGPT and AI products.  If the people who are creating these software entities are doing a half-baked rushed unvetted job then what do you think the end user will wind up doing/creating.  I think this is going to be a case of "Who Watches the Watchmen?".  Once these AI driven products are released to the masses and wide spread, what's to stop unvetted information.  Information doesn't have to be blatantly wrong to be harmful, it can just be "kinda" wrong and do massive amounts of damage.  This is my point.  Once we as a society are reliant upon AI generated services that are true/correct enough to be believable we are doomed.  It will mostly affect those who are underserved and disenfranchised first.  I would imagine that some of the readers of this article won't be able to relate, but eventually, once the lives of the people that the upper social economic classes are built upon crumble, so will their livelihood as well.  

"I Ain't New to Dis!"

 Yes, I said it, ebonics and all.  I am not new to building robots and making stuff.  I've been pursuing engineering/mechanical/industrial design since elementary school.  Ever since I was introduced to giant fighting robots and the likes from all the 70's/80's Japanimations/cartoons that proliferated kids programming on television.  I could not get enough of it and I determined within myself to figure out a way to build what I saw on the screen.  Now here I am in 2023 watching all of the amazing robots that have been designed by so many smart people and institutions, and I'm like God, what have I produced?  Why am I not, at the least, involved in some innovative robot project?  Why don't I ever see faces that look like mine on the teams that build such projects?  Why don't I have any friends that build bots?  

I see all the young people with access to this new influx of technology, 3d printers, laser cutters, cnc machines (3 to 5 axis) and so on.  I came up before this era, where I had to hone my model making skills, sketching, and soldering.  Now it's time to add the old to the new and get back into robot making.  

Just to give a little taste of my previous projects.  I'd like to introduce you to my Squatter Buster. 

This little unit was the property of the Cobb County police dept.  It originally came to me as a Traxxas E-Maxx and a old paintball gun.  My task was to get it all assembled together with a camera/transmitter and make it into a radio controlled less lethal deterrent.   Using an analog camera and video transmitter along with a JR airplane radio, the Badger was born.  Here's a run down of its features:

>Servo controlled Pan/Tilt

>Full auto fire (switch activated)

>Retained stock speed... (too fast to drive indoors lol)

This would be my first foray into this type of platform, but it wouldn't be my last.  Here's a sample of it in action evicting a squatter (me) in an   empty house.  Enjoy!  Be sure to add your email to the blog so you won't miss all the robot/mechanical goodness that is coming.

 

TRS/Sci-Bot Worx - Waste Not, Want Not ep1

 Practice what you preach is what I try to live by and I've been screaming from the mountain tops that mankind has become too wasteful.  I believe that much of it is driven by the over supply of goods in attempt to get a return on investment and corporate greed.  Not here at TRS/Sci-Bot Worx!  Nothing breaks for good here.  We go to great lengths to either repair an item that is broken, reassign that item for another purpose, or dismantle it and reuse its components.  Today we will be demonstrating this first hand.  My wet/dry vacuum from 2004 has run into an issue.  The original aluminum fan blade has developed cracks in it due to stress caused by the use of the shop vac as a leaf blower.  The gyroscopic forces generated by the motor and the moment forces caused by the movement of my hand while blowing leaves have proven to be too much for the OEM blade.  

So upon this discovery, instead of throwing in the towel and spending another $139 for a comparable shop vac. I decided to 3d print a new fan blade (that was the easy part).  

Upon further examination of the motor shaft/axle, there appeared to be a large washer at the base that the previous fan blade sat against.  I thought that maybe this had fused to the motor shaft over time because I couldn't fathom how the motor could be assembled with this washer welded on.  So I tried to remove it with plyers, then tried to dremel it off all to no avail.  Some how it was apart of the motor shaft!  The next course of action was to see if I could get it mounted in the lathe so I could clean up the shaft that way.    

Seeing that I couldn't remove the motor from the plastic housing it was in, I knew that it was going to require that I get really creative in order to get in on the lathe.  I thank God that It just barely fit into my chuck and I was able to use a live center to support the free end.  This enabled me to turn what was left after I had hacked most of the washer off with the dremel.  

So back to printing.  I quickly modeled up a fan blade using the original aluminum one as a reference.  Then sent the file to be printed on my modified Ender 7, which made short work of it, and I mounted the blade to the motor shaft.  Mission accomplished, so I thought.  I reassembled the motor housing and flicked on the power switch.  As I had hoped, the vacuum began to suck again.  I actually thought the the suction had increased (I was wrong), but it was good enough for me to return to using the vac.  All of the celebration came to an end though once I began to use the vac as a blower again.  It was short lived as the motor began to heat up, the PLA fan blade began to warp causing it to rub against the motor housing.  

Not being one to give up easily, I then formulated a plan to create another fan blade, but this time it would be printed out of PA-12 CF (a Nylon with chopped carbon fiber filament).  Thankfully, I was able to use my recently finished custom drying oven to dry the PA-12 CF so that I could print the new blade.  The blade came out prefect, expect for one small detail.  Upon revisiting the OG aluminum blade I realized that I had the vanes of the fan blade curving in the wrong direction.  This required me to correct the model and then make another print.  Thankfully all that took place without a hitch.  

This little project was so much fun (not really, I just need my vacuum to work) that I even made a little video of the repair down below.  Thanks for reading and see you on the next repair.

Boston Dynamics' Stretch at MODEX!

 I had a chance to attend the MODEX show this week and was pleased to see that Boston Dynamics was in attendance with their Stretch platform on full display.  I love seeing robotic products going from the prototype (work in progress) stage to the production model phase and boy is this thing a beauty.  The gray and white color choice is clean and non-threatening.  The simple geometries of the turret base and the main body are simple yet elegant.  I also love the wide and flat appearance of the turret mast that houses the vision systems.  All in all, I'm very impressed at the way the design came out and would enjoy picking the brain of the head designer(s) that produced it.  Unfortunately, I wasn't able to stay at the Boston Dynamics booth for long and I did not take enough pictures.  Here are the two shots that I snapped and you can watch an older video of Stretch in action below. 

Hopefully I'll be able to share some more goodies from the MODEX trip.  Stay tuned!

Never use gloves with rotating equipment - PSA

    Well, I was told this before and I agreed with the statement, but it's not until now that I'm determined to implement it.  You see, during my past employment, I've worked along side several very skill machinists and from time to time they would impart nuggets of wisdom to me.  One of those nuggets was not to wear gloves while using a lathe.  I tend to wear gloves often while working in the shop and I normally don't remove them when I use the milling machine or lathe.  Usually, I'm aware that I have them on and keep my hands away from anything spinning, but last week I made a careless mistake.  I was in a hurry to get a freshly printed part installed and needed to drill some holes quickly.  No big deal, right? As I was drilling the holes, I decided that it was a good idea to brush the chips, that were being created, away with my gloved hand.  I did it twice.  The first time, no problem.  On the second time, the drill bit snagged the glove at the tip of my pointer finger and began to wind it up.  It quickly pulled my finger and therefore my hand into the bit.  I immediately began to pull away, but I could not free myself.  The glove ripped and the digit of the glove became like a tourniquet, tightening more and more as each second went by.  I grunted in panic, eventually realizing that I could not pull away, I tried to turn off the mill.  I overturned the three position switch, making it temporarily go in reverse and then I finally switch it to the middle, cutting the mill off.  At this point the drill bit stopped and I was able to remove the glove and therefore my hand from the mill.  There was no blood just soreness.  Thank the Lord that the glove took all of the cutting from the edge of the drill bit.  I am so thankful that there wasn't any permanent damage done and that this serves as a reminder to adhere to safe practices around machinery.  

Shower valve repair - The Engineer vs. Life

Well let me tell you what happened on yesterday. The shower head in the master bathroom has been steady dripping over the last month or so and I have been planning to set aside some time to repair it.

Well life determined that I would be repairing it sooner rather than later, because the drip almost became a steady stream the night before.  Thankfully, I had done my research on the valve so I knew what to get and where to get it.  I also knew that there was a high probability that the valve would be stuck in the pipe requiring the use of a valve puller tool to get it out.  

Upon purchasing the replacement valve I saw that they included  a plastic twisting adapter that was to be used to "twist" the valve out of the pipe. After several hopeless attempts it had become apparent that this adapter was useless in this situation, I needed a valve puller.

The valve puller was only $15 at the local hardware store, but me being the Engineer type that I am, I thought, "surely I could come up with something that would work".  So after pondering on exactly how the valve puller works, I began to search the shop for something that could fit the bill. Eureka!  I remembered that I have a set of hole saws in my shop and they have a mandrel.  Surely there is some way to adapt the mandrel so that it will operate as a valve puller, I thought.  So I conducted a mini scavenger hunt around my shop to find what parts I would need.  First, I needed a screw to be inserted into the stem of the valve.  Also this screw would need to be able to attach to the mandrel in some fashion.  Secondly, I needed some material that the mandrel would be able to push against that had a large enough diameter to be able to be braced against the pipe the valves was in.  Lastly, I needed some sort of handle in order to hold the make shift valve puller when I go to rotate the knurl wheel.   Thankfully I found a 10-32 socket head screw, an old Delrin wheel hub that I had made, and an adjustable wrench.

To make this all work I needed to turn the head of the screw in order to allow it to be inserted into the end of the mandrel.  I needed to lop off the large diameter portion on the delrin hub, and I also needed to increase the size of the through hole in the delrin hub.  So as I began the work, the first stop was the bench lathe.  Thankfully, I was able to make short work of turning down the screw's head diameter.  Next I needed to lop off the large diameter on the delrin hub adapter.  The cut off tool that I was using wasn't quite set up long enough to reach the inner bore on the delrin without crashing into the workpiece, so I had to improvise with a hacksaw.  After making that cut, I cleaned up the face of the delrin and increased the inner bore diameter to go over the valve stem.  

At this point, I'm ready to give my mandrel valve puller a shot at the stuck valve.  So I go to the shower and install the delrin mandrel combo, add an adjustable wrench and plyers, and start turning the knurled ring.  So far things are looking okay.  The pins in the mandrel are pushing on the delrin and the delrin is pushing on the pipe, but then I come to the end of the screw travel on the mandrel and the valve is still stuck!  Well it turns out that the delrin is pushing on the pipe and on the valve at the same time.  I come to the realization that there is supposed to be a notch in the valve facing end of the delrin so that it doesn't press against the valve while it's pushing against the pipe.  Back to the shop to make a quick notch with the milling machine.  I found the closest end mill to 9/16" and made a cut about 1/2" deep through the approximate center of the delrin part.  Now quickly back to the shower for the second attempt and voila!  This time when I start to turn the knurled ring I see the valve start to move.  I turn the ring until It's gone the full travel on the mandrel and then try to pull out the once stuck valve.  Success!  The valve comes out and I now have access to the pipe. 

The replacement of the valve went smooth from this point on and there is no more drippy shower head!  Would I recommend doing it this way?  No, not unless you just happen to already have these tools at your disposal or can not locate the valve puller from your local hardware store.  With my time, tools, and materials cost, I've greatly exceeded the $16 that was being charged for the valve puller at the store, but it's good to know what can be accomplished with the resources you already have at home.  

XBOX Design Lab inspired R/C Transmitter

 This has been forever in the making.  I'm in the process of designing and building my first 15 - 17 channel RC radio transmitter! I've always been fascinated with all types of control surfaces, but specifically remote control car controllers.  I have a growing assortment of all types of radio controls and video game controllers.  Now, I'm combining what I see as the best of both worlds into my very own personal radio.  Be sure to tune into my new livestream where I'm designing the radio in real time and going over the why and the how of what I'm doing.