Babysitting ChatGPT

-

 After my initial excitement with chatGPT, I have been trying to use it in the real world. I needed to plan a certain programming task and I was expecting to get it done quickly and easily with the help of AI, but the result has been quite different. 

The first big problem is that code tends to get cut with no warning sign. You get a snippet of code where part of it is missing. Sometimes it is more evident than others and when you mention that in the chat, the machine goes back to produce another snippet that many times is cut again. Not so smart or efficient. The only workaround I have found digging in the OpenAI discourse forum is to report the last line of the output so the machine can continue with (hopefully) the rest of the code. That approach seems to work better as I guess the culprit is the limitation of the maximum length of the output code.

But the next problem is the code generated did not work. Sometimes the reason was quite evident like they are trying to use a data type that does not fit the needs of the object or method it is applied to (like an unhashable type for a dictionary key). Other times, there is not a proper match between the parameters passed to a method. The bottom line is that once you manage to get the complete output created by the AI, something may not work properly.

On the other hand, while you can guide the code creation with your criticism and suggestions, the process ends up being frustrating at times, where you end up running circles and getting this feeling you have a willing but not very sharp helper. 

And of course, all these interactions require time and effort on your side. A time that you are taking away from your own work of actually solving the problem. I felt like I was wasting my time trying to explain to a not-so-bright machine what I wanted to end up with non-working code that now I have to figure out how to fix.

So I think I am going to wait another five years to see if the new AI does significantly better. The current "help" I am getting is not really helping me. After all, I guess these companies are not in the business of helping me for free, but more in the process of getting help from their users of free tools to fine-tune their AI systems. So for now, thanks but no thanks.

Comments

Post a Comment

Popular posts from this blog

VFD control with Arduino using RS485 link

-
Image
Variable Frequency Drives ( VFD ) allow the control of spindles so speed can accurately be controlled and a detailed acceleration profile for the spindle and reverse rotation can all be handled. In essence a VFD is an three phase inverter for three-phase AC motor. I am using a popular (I mean cheap) Chinese VFD and though the reference manual is not great, I could see there is a built-in RS-485 port. I usually control the start, stop and speed selection using the keyboard on the unit but I thought it will be more useful if I could control everything from the same Arduino is doing our CNC table control. Some cheap RS-485 off eBay and some lines of code later, I can start, stop and change the speed from an Arduino. What a cool thing to have! Some configuration of your VFD are needed before you can use it like that.  You need to set PD163=1 (I am using address 1 in the code). PD=164=1 (for setting serial to 9600bps) and PD165=0 (for using ASCII and 8N1 character format)
Read more

How to get sinusoidal s-curve for a stepper motor

-
Image
After some experimentation, and delving a bit into the math of motion, I realized there can be a significant difference on the motion of a machine by just making small changes. A common and popular approach to smooth one axis of motion is to use a trapezoidal speed pattern, where acceleration and deceleration are uniformly accelerated movements (aka constant acceleration motion) separated by a portion happening at cruising speed. The problem comes with the fact that, while that approach is simple to understand and to code, it does contain sudden changes in the acceleration, that show as spikes on the acceleration derivative, called jerk. These sudden changes in the acceleration translate into undesired vibration in our machines. Most of that can be eliminated by selecting a motion pattern that does not contain sudden changes of acceleration. If we chose a mathematical function for our speed whose second derivative is continuous, then we reduce system vibration quite a lot. One ma
Read more

Stepper motor step signal timing calculation

-
Image
While we use stepper motors in many devices, the timing of the step signals is not always well understood. Most of us can see how a fixed speed is a bad idea unless the speed of motion is really slow, but it is fixed speed what is really to generate with simple code. You just can use the Arduino Blink example to send a fixed frequency of pulses to a stepper motor driver. If too slow, just reduce the delay value. But for most real-life applications variable speed controls will perform better. These systems will speed up the motor to a given maximum speed and it will slow it down to stop at the desired step count. But understanding the math behind this process is sometimes not obvious. Let us start with the simple case of the so-called trapezoidal speed motion profile. Here motion takes place in three phases: acceleration, constant speed (cruising) and deceleration. But the point is how can our code create such a behavior. Stepper motor drivers usually accept two signals to control
Read more