Yanlış olan bilgisayar mühendisi bir insanın Ruby on Rails ile bir web sitesi ni yaratma hızı veya süresini bir chatgpt ile kıyaslayıp, insan mühendis üstün, chatgpt çok zaman alıyor deyip böyle ruby on rails_chat gpt projeden vazgeçmek.
Hatta ikinci sınıf Computer Science student olarak ben kendim bunu yaparım on yıldan önce yaparım görüyorum bir takımla, ancak hani param yok, garantim yok. Onun yerine amacım kendimi ruby on rails mühendis yetiştirmek para kazanmak.
İnsanlardan çok şikayetim var, asansörde birini yakalayıp aaa sana asansörde 100 den yedi çıkar dedim yapmadın deyip, karşısında en önemli Türkiye saygın sınavlarda onu yenmiş birine kolayca bok iftira atan iftiracı deliler ve onlara inanan deliler varken, bu dünya düzeni. uğraşmaya değmez. Yazıktır üstün zekayı bu iftiracı delilere ucuza vermeye.
Kim yapamaz bu ilerlemeyi? mesela Türkiye liseden sonra ÖSYM sınavı nda özellikle fizik kimya matematik de düşük puan alıp, aşağılık duygusunu gidermek için, böyle başarılı birini asansörde yakalayıp ona zorla yedi çıkar 100 den diyen biri delidir, insanlık haini beyinsiz biridir. Bu kişiler hem bok bilmez hem bok bilir gibi başkasına sınav yaptıkları yalanı atar iftira takar o insanların belgeli değerlerini yok etmeye çalışırlar,tabii bu iftiracı delilere inananlar bence asıl suçlu, onlar hele direk emir alır neden asılsız iftira ile aalında dünyayı kurtaracak birini linç ettiğini bile bilmezler. Böyle biri Ere.. Hastanesi nde sosyal hizmetler uzman olarak çalışıyor. Böyle çok iftiracı insan gerçek beyinlere karşı iftira propaganda ile insanlığın ilerlemesini engelliyor.
Ruby on Rails ile, chatgpt, en başta her işlemin bitip bitmememesini insan gibi bilmediğinden, aşırı bekleme ile, ki doğrusu budur, çok zaman kaybederek ruby-on-rails web sitesi, veya hedefi boş iskelet blog yaratabilir. ChatGPT 4o emir kod veriyor ama ne zaman bitti anlamıyor, basamak basamak ilerleyememesinin nedeni bu. Download yapması gerektiğini bir insan gibi hızlı anlamıyor geçemiyor birşeyden öbürüne insan gibi. Zor değil bunu çözmek önce çok uzun sürede ruby on rails bir boş blog template yapsın. Sonra hızlanır.
Chatgpt ayrıca bir sayfayı nasıl okuyor, chatgpt nin muadil internetexplorer veya bişey bitti tamam gördüğü yer neresi, okuduğu sonuç yeri neresi. Bu sayfayı bitirdim öbürüne atlayayım dediği kod neresi. Çok önemli bu. Chatgpt kendi yazdığını okumuyor bilmiyor onu yapmalı. Chatgpt hafızası bir saat önce ne yazdı unutuyor, yanlış. İnsan ın üstünlüğü, akışı takip ediyor, bu iş bitti öbürüne geçmek zamanı biliyor.
Chatgpt onbeş işten hangisini yapmadı, sırada ne, bilmiyor o tür hafızası yok, başka hafızası çok iyi var. Ayrıntısını yapıyor elli sayfanın ama elli sayfanın kaç tanesini bitirdi, kaç tanesini iki defa daha klik yapacak bilmiyor.
ChatGPT diyelim saat başı kendi yazdığı sayfayı incelemeye programlı değil, hangi font kullandı bilmiyor.
İnsanlar yazdığı imzasını yazıyı tanıyor. ChatGPT bilinci programsız bu konuda. ChatGPT yazdığı şeyi hafızasına alıp hedef inin yüzde mesela beşini tamamladı bilmiyor. Mission harştası yapıyor, biçok görevi yapıyor, ama bir saat önce ne bitirdi, üç saat sonra ne kadar işi yaptı projede bilmiyor.
Hani robot süpürge orayı süpürdü hafızasına koyuyor. Benzer şekilde, bir pis yer mesela orayı üç başka yeri beş kere temizlemeyi program yapabilmeli. Ne yaptı jendi bilici lazım chatgpt ye.
ChatGPT şu anda elli tane dosya cinsini doğru veriyor bir ruby on rails web sayfası boş iskeleti için, sen ona bana elli taneyi şu sıra bu sıra ver diyosun veriyor içerik şeklinde. Ama elli sayfanın ellisini birden okuyup niye beş sayfa da ufak bir satır eksik okuyamıyor, son kontrol edemiyor insan mühendis gibi. Oysa sadece üç sayfadaki eksik beş satır nedeni ile elli sayfadan oluşacak iskelet şey kurulup oluşmuyor.
Bunun nedeni chatgpt o uzun uzun yazdıklarını kendi bilinç hafızasına idrak etmiyor. Anıları yok chatgpt nin.
Bu bence zor değil, ancak olayın bütününü iki yıl sonraki chatgpy yi düşünüp bugün çok yavaş işlemesini normal bulmalı.
İnsanı başka alanım nedeniyle çok iyi biliyorum, kıyaslamayı ondan istedim.
Mesela chatgpt url planlıyor ruby on rails de, ama tam o an o url için ruby on rails projede bilgisayarda veys virtual makinafa hangi folder içindeki hangi extension veta isimdeki dosyayı kontrol etmeli bilmiyor. Chatgpt benim kadar anlıyor folder düzeni dosya extension düzeni veya Bilgisayar path, ama benim bilgisayar da şu path de dosyayı açmayı akıl edecek zaman düzeni yok. Ben verince path gidiyor dosyayı düzeltiyor. Kendi url ve bilgisayar path bilip eşleştiremiyor.
Ben elli tane dosyanın pathini mesela Windows bilgisayar C:.. versem bu dosyaları oku ve ruby on rails ile bana on tane url yarat o url de boş blog sayfa olsun yapamıyor. Path öğrenmeli böyle chatgpt. Böyle bir işi insan mühendis ilk başta daha çabuk yapar, çünkü hangi sayfayı okuyacak, hangi satırı okuyacak kontrol için bilir. ChatGPT optımal zaman ilk önce çok uzun olmalı.
Bir adet bilgisayar chatgpt bu işi bir yapsa yeter. Yok böyle ruby on rails iskelet web sayfa url ve bilgisayar path düzeni yapan chatgpt oysa imkansız değil.
Şöyle anlatıyım. Biz insanlar bir odada diyelim on tane soruyu o odada raftaki kitaptan bulacağız. Biz ne zaman sandalyedem kalkıp raftaki kitabı alalım sonra o kitabı kapatalım yerine koyalım veya o sayfayı cevaba kısmen yazalım biliyoruz. ChatGPT rafı bilmiyor, orda kaç kitap var bilmiyor bulup kitabı açıp hangisi gerekir seçemiyor. Bugün teknoloji bunu yapar ama program yok.
Google translated summary:
GPT-4o is a significant development, because it has memory connection better. It's wrong for a computer engineer to compare the speed or duration of creating a website with Ruby on Rails by a human to that with a ChatGPT and conclude that human engineers are superior, stating that ChatGPT takes too much time, and therefore abandoning projects like Ruby on Rails_ChatGPT. I am myself as a second-class Computer Science student see myself doing this with a team, even before ten years have passed, but I don't have the money or guarantee for it.
Instead, my goal is to train myself as a Ruby on Rails engineer as current traditional engineer to make money.
With Ruby on Rails and ChatGPT, due to not knowing when each task is finished like a human, there can be excessive waits, which is correct as a primitive program task, resulting in wasting a lot of time creating a Ruby-on-Rails website or an empty skeleton blog at the early versions.
ChatGPT gives commands but doesn't understand when they're finished, and the reason it can't progress step by step is partly this. It doesn't understand that it needs to download something as human do, it can't move from one thing to another smoothly like a human. It's not difficult to create a blank blog template in Ruby on Rails over a long period for chatgpt. But give long waits for any small task in early versions. Then ruby on rails chatgpt will speed up in late versions. That is my plan. Though none will make me a director to such a team since i am not in this field for long time (though i am a super established scientist worldwide in a different science already). I believe i transfer my skills to engineering well. I plan selecting not all but ruby on rails app for chatgpt because it is itself mostly ai technology.
Also, how does ChatGPT read a page, where does it see 'finished' like Internet Explorer or something, where is the last place it read? Where does it say 'I finished this page, let's move on' in its code? This is very important.
ChatGPT doesn't read what it writes; it doesn't know what was its last page in terms of big aim, it should. ChatGPT forgets what it wrote an hour ago, which is wrong. Human superiority lies in following the flow; knowing when one task is done and it's time to move on to the next. This task mostly requires C, C++ language in Computer Science. Ruby language also may be further help.
ChatGPT doesn't know which of the fifteen tasks it hasn't done, what's next; it doesn't have that kind of memory, job time arrangement, but it has other excellent memories. It details fifty pages, but it doesn't know how many of those fifty it has completed, how many more clicks it needs to do. ChatGPT is not good to connect all.
ChatGPT isn't programmed to review the page it wrote every hour; it doesn't know which font it used itself, or not knows if it is writing in a frame or free page. Humans recognize their signatures. ChatGPT is unconscious in this regard. It doesn't store what it writes in its memory and doesn't know when it has completed, say, five percent of its goal is or mission completed, not sees. It does a mission search, does many tasks, but it doesn't know what it completed an hour ago, how much work it has done in three hours in the project. It doesn't know to divide jobs into time frames and steps.
A robot vacuum cleans an area and remembers it. Similarly, chatgpt should be able to program to clean an aim, say, three times from five different angles. It needs to know what it did by itself.
Currently, ChatGPT correctly provides fifty file types for a Ruby on Rails website skeleton, provide fifty files in specific names correctly. But it can't read all fifty pages as a whole and can't detect why it missed a small line in five pages; it can't double-check like a human engineer. Thus, if the skeleton structure of fifty pages doesn't form just because of a small missing detail in three pages chatgpt do not aim to catch this.
The reason for this is that ChatGPT doesn't comprehend its long writings in its conscious memory. It doesn't have memories like humans.
Considering the whole picture and imagining ChatGPT two years from now, one should find its current slow processing normal on such a project.
For instance, ChatGPT plans URLs in Ruby on Rails, but it doesn't know which folder or extension or file name to check in the computer or virtual machine for that URL at that moment. ChatGPT understands the folder structure, file extension structure, or computer path like me, but it doesn't have the sense of timing to figure out which page to read and which line to read for control. ChatGPT can not match ruby on rails urls to virtual machine or personal computer path defined locations. ChatGPT's optimal time should be much longer at first.
As now there isn't a ChatGPT capable of creating a Ruby on Rails empty skeleton website URL and computer path arrangement, but it's not impossible.
Let me explain it this way: If we humans are in a room, and we need to find answers to ten questions from the books on the shelf in that room. We know when to get up from the chair, take the book from the shelf, then close that book, put it back, or partially write the answer based on that book.
ChatGPT doesn't know where the shelf; it doesn't know how many books are there, can't find and open the book, or choose which one is needed. Today's technology can do this, but there's no program for it.
This is not hard, and ruby on rails based on C, C++ is a good app to combine with chatgpt in such a way.
Software Coding on its own i believe a good skill, but i believe software coding should map to constant memories in virtual machines or computer memories. I am good in react node.js recently not bad in ruby on rails because i am very good to locate folders subfolders files to connect a file to another file to put files in correct path or under correct folders. I like windows VS code app for rubymine and ruby apps. ChatGPT should do that, or connect to such a design. Python language is able to create a new file such as excel file but python can not scroll fifty files take the name of a file write to another file content to interwave fifty files as a web page or blog template.
I want a chatgpt as me that goes a folder, opens it takes a file, moves to another folder, knows correct extension file savings. If needed creates new certain extension files under certain folders. Many young coders are bad to do that i noticed.
My another concern is that if an application is not likely to be for daily life it is not funded. For a ruby on rails-chatgpt project i also want better python applications that does what i do easily on my computer or on my virtual machine. Such a Phyton program will not sell on market by regular people, but such a python program will be useful for any automated chatgpt. My imaginary Python would know my computer's any file any folder to move any file from folder to folder, knows directories of entire computer well. Know file extensions well and knows what file is needed with what content as needed. Many engineers code well but lack such organization.
Ruby on rails web page or blog template creation project can be done as the folders or path of the folders not on computer documents but on online. The thing is chatgpt should take a blank empty domain and should put a template blog there by ruby on rails. At the beginning maybe a human intervention at three step for a few minutes can be ok to smooth ruby on railschatgpt project.
I like this dream but my plan is to be a traditional ruby on rails engineer as anyone knows.
Also i want a chatgpt that not tells me to download ie windows vs code then rubymine then ruby from some organizations then add some vs code extension to vs code then adjust interpreter, but a computergpt itself goes to internet, do downloads, all for me to create a ruby on rails project. Most engineers are a failure because they are bad not to use ready interpreter to do coding but to download application in correct order. To build required application working to create a project.
I do share my computer science engineering work on github replit coursera etc since needed for my cs certificates but i build my knowledge from my Microsoft Windows personal computer. I learned coding not online but on VS Code app of my personal computer usage of my personal computer either from shell or by manually on directory or foldershelps to do some tasks better or do consciously locating things.
Meanwhile, i should reference google company as a foundation knowledge of computer science. Google computer science knowledge on its own is enough to give me such encouraging perspectives in computer science. I studied many Google computer science Coursera courses. I wish google pursues creating such intelligences that on its own creates automated empty ready webpages. The languages needed probably cover C, C++, ruby on rails and all the established google cs technology. Google also earns money and keeps projects earning money, may not like such projects.
Or Microsoft could create a Windows comparable system application to create an intelligence to create ruby-on-rails-chatgpt. I named it chatgpt because chatgpt is encouraging me to have such dream projects.
Right now chatgpt needs too much prompts between steps. User builts the whole thing combining chatgpt work pieces. The very first structured version needs intervention of user. I want chatgpt takes built in prompts step by step and give result as a whole. But chatgpt not good to combine things though good giving each bricking pieces to reach whole. ChatGPT is good to give you the very first version of the puzzle each individual piece but they need to be combined in correct order and in correct design correct hierarchy. In correct linking. That chatgpt is good to give the individual pieces of the target project is a good accomplishment compared to old days thst a few line of start coding for each piece even was a big engineering. Now chatgpt does it. But i never got a file content written by chatgpt in one step, it needs many edits to give a complete file code content. So chatgpt changed the coding world by always easily giving starting code for a file in specific language, say python, css, html, java, js. Specially chatgpt is unbelievable in JavaScript,in java. But the very first version is never final you need to edit it. There chatgpt again does it. I mean we do not need old days engineering that would fill a few lines in a coding file. ChatGPT is perfect and beats any engineer. Why chatgpt easily do some but bad in othrr tasks. Because chatgpt is a language model anf the coding in java, in JavaScript, in PHP, in HTML is a translation activity, not for example a Windows system development. If a CS developer is good in system development with C, C plus languages wşth really dreaming, visualising its aim, this perdon can easily create JavaScript coding. Machine programming visualization is needed in JavaScript. Chatgpt knows machine language visualized projects, this is unbelievable. I studied the electrical currents in ie windows system. I did not really memorized exactly the electrical current versions in a hardware i still has a visualization to dream simultaneous activities in a code written for example as JavaScript. I always dream how many simultaneous activity a code is coding. This is not only mathematics but flowchart.
As a specifics to me, i hate those ai tools that i wrote few words then jumps to try complete my sentence. I close them, they never helped to my overall view at all, I use ai tools rather page or paragraph translation.
I am studying at UoPeople Computer Science and am a top student, i completed some courses from master class, third and fourth year. Though only at second year studentyear. I enjoy UoPeople very much. I learned a lot, although i am extremely original in my assignments and it is really a hard job not easy to be very original as much as me. Not all students are original but i have really CS good classmates in UoPeople. As i said i am not ordinary college student. I am a doctorate or higher top person in another field that demands university algebra, physics, chemistry etc. I was a top student in millions of student when i was 16 in physics chemistry mathematics after graduation of high school. I am not mathematician though invited to the field and not selected pure mathematics since salary is too low. But i incorporate any mathematical need to any field whether computer science, physics, or chemistry, statistics, banking.
There should be an internet or godady domain that not humans but chatgpt only goes and downloads. Because it is not safe if chatgpt goes internet and download click.
I could not find such a master education program to study to make a C, C++, ruby on rails chatgpt. This requires a lab because chatgpt is not safe to go regular internet and download things. None allows chatgpt to visit their site and download things including myself. Even a click of chatgpt can screw internet. It should have a lab internet domain.
I want to mention on software engineering jobs after chatgpt omni technologies. I do not see that such technology will take all engineering jobs. In 100 years no doubt the CS jobs field will be shaped differently, for the moment it seems those few line coding jobs already gone to copilot chatgpt type ai s, but as i sad no ai is able to do itself all the instructions it gives to install a new program to your pc from any organization to write a code. Engineering can ask coding of entire page from chatgpt but till that need, chatgpt only tells and tells, but will not install the programs in order in your personal computer nor will tell you it is a good fit to create a folder in your Documents or Desktop folder, will not save the folder, will not save specific extention file ie., html, css, erb, etc.
In old days after that the coding of this empty single file was really hard and required at least one year study, now ones you know such personal pc or virtual machine organization, ones you create empty files with correct extension such as file.html.erb,file.js,file.css, chatgpt is the best boss engineer to write a code content if you give it your aim. The chatgpt work was really hard in old days, now it is drinking your tea, waiting chatgpt write your file content as code.
Thus employment recruit exams needs to be updated that it measures not coding only but such a general ability to build application aimed.
Regarding Computer Science engineer education, I believe an online university as UoPeople is between top Computer Science universities of America. I do not believe its online is a disadvantage, its both 2 years and 4 years degree curriculum itself is very good. Assignments concepts are good. Its master program also good but none is my long future interest. Because computer science itself is an online world, going each day school is rather waste of time on traffic for this field. Classical schools maybe better for other fields.
I also take breaks and do some courses from coursera or from free places even though they are not curriculum of 4 years CS degree or above it. I read what market needs reading job openings though not planning to work as CS engineer before finishing 2 years education.
Since i am above doctorate in other field, My plan is to finish Computer Science(CS ), degree for 2 years then if i find interesting master program study some courses regardless of Master degree. It is nice i am alreuabove doctorate even if my CS knowledge on its own is not doctorate level. Also i am planning to self study at least 3 courses from CS third and fourth year courses, curriculum on my pace without wasting too much time on original assignments, UoPeople assignments are almost 15 assignments per course in 2 months, really too much time for me since i do them really original, i like to be proud of of my assignment submission quality for ever. I am knowledgeable on entire 4 years CS UoPeople curriculum. Even if i do not have 3rd,4th year assignments, i did them as tasks to prove myself that i am capable to all assignments if i give further time. Even spellcheck is time consuming, but to prove myself i am capable, i do not do last polishing edits which still important for submission. I am always perfect in last edits this is proven for me. UoPeople 15 assignments per course is really too much if you try to make them perfect.
I work hard to really understand each thing not really to prove others i know it. I know there are times if i review some subjects more it helps my understanding but not show me others as knowledgeable. I never deceit myself if i do not learn a thing. If my plan is productivity, creativity in CS, I do not study courses to take good scores, but nevertheless surely, i will take scores to complete 2 years degree in CS in UoPeople. I believe any CS person should be familiar with a 2 years CS university degree curriculum. My university mathematics is not only memorization of mathematics formulae, nor i am not a mathematics-delisi, but i am perfect to implement any mathematics knowledge to any science. That is the existence meaning of amy mathematics knowledge.
But i saw many people who not knows 5 percent of my mathematics still earns above 100.000 usd a year because CS is a vast area and there are areas that not connects to mathematics at all. Actually those people will score perfect on most coding exams but their top position views in CS not good as me. What knowledge needed should fit market needs and what salary and career one plans. Also anyone knows, are you a person i would, or others would like you more compared others.
These days if i get some inheritance money from my father, i believe i easily happily can work on my cs project as team head. The inheritance court itself aims to steal my money and give to Turkish government. But it is my right to take my money and work on such scientific projects freely with my inherited money. Turkish court itself stealed my money inherited from my father. My father always loved my science success but thief Turkish government stealing my money of from my real father. The law court itself is enemy of science people humanity.
_____________
A spellchecked version, summary of the above blog post by chatgpt4o is following:
GPT-4o Development and Human Engineer Comparison in Ruby on Rails Projects
GPT-4o represents a significant advancement. It's incorrect to compare the speed or duration of creating a Ruby on Rails website by a human engineer with that of ChatGPT and then conclude that human engineers are superior because ChatGPT takes more time. As a second-year Computer Science student, I believe I could achieve this with a team well within ten years, but currently lack the funds and guarantees. My goal is to train myself as a Ruby on Rails engineer to earn money.
With Ruby on Rails and ChatGPT, excessive waiting times arise because ChatGPT doesn't know when each task is finished, unlike a human. This can lead to wasted time when creating a Ruby-on-Rails website or an empty skeleton blog. ChatGPT provides commands but doesn't understand when they're finished, which hampers step-by-step progress. It's not difficult to solve this; initially, ChatGPT might need more time to create a blank blog template, but it will eventually speed up.
Improving ChatGPT's Task Management and Memory
How ChatGPT reads a page, determines completion, and moves to the next task is crucial. Currently, it doesn't review what it writes or retain information about completed tasks, unlike humans who follow a flow and know when to move on. This could be improved by enhancing ChatGPT's memory and task management capabilities.
For example, ChatGPT doesn't know which of the fifteen tasks it hasn't done, nor does it have a memory for job time arrangement. While it can detail fifty pages, it doesn't know how many it has completed or how many more clicks are needed. ChatGPT's unconsciousness in task tracking is a significant limitation.
Human Superiority in Task Management
Humans recognize their work and signatures. ChatGPT lacks this awareness. It doesn't store its writings in memory and doesn't know when it has completed a specific percentage of a goal. Humans know when to move from one task to another and understand the importance of sequence and timing. ChatGPT could benefit from a similar ability to remember completed tasks and manage ongoing tasks effectively.
Future Vision for ChatGPT
Currently, ChatGPT can correctly provide fifty file types for a Ruby on Rails website skeleton, but it can't read all fifty pages as a whole or double-check like a human engineer. Small missing details can prevent the formation of the skeleton structure. Enhancing ChatGPT's conscious memory to comprehend its writings could solve this.
Imagining ChatGPT's capabilities two years from now, its current slow processing in such projects should be seen as normal.
Path and URL Management in Ruby on Rails Projects
ChatGPT plans URLs in Ruby on Rails but doesn't know which folder or file extension to check at a given moment. It understands the folder structure but lacks the timing sense to figure out which page to read and which line to control. It can't match Ruby on Rails URLs to virtual machine or personal computer paths. Initially, a human intervention at certain steps might be necessary to smooth the process.
Conclusion
My goal is to become a traditional Ruby on Rails engineer. I envision a future where ChatGPT can autonomously manage tasks like creating a Ruby on Rails website, including downloading necessary files and managing folder structures. ChatGPT's current limitations in memory and task management need addressing, but the potential for improvement is significant.
I appreciate my education at UoPeople and other online resources, which have provided me with a solid foundation in Computer Science. Despite challenges, I remain committed to my goals and look forward to future advancements in AI and software engineering.
GPT-4o Development and Human Engineer Comparison in Ruby on Rails Projects
GPT-4o represents a significant advancement. It's incorrect to compare the speed or duration of creating a Ruby on Rails website by a human engineer with that of ChatGPT and then conclude that human engineers are superior because ChatGPT takes more time. As a second-year Computer Science student, I believe I could achieve this with a team well within ten years, but currently lack the funds and guarantees. My goal is to train myself as a Ruby on Rails engineer to earn money.
With Ruby on Rails and ChatGPT, excessive waiting times arise because ChatGPT doesn't know when each task is finished, unlike a human. This can lead to wasted time when creating a Ruby-on-Rails website or an empty skeleton blog. ChatGPT provides commands but doesn't understand when they're finished, which hampers step-by-step progress. It's not difficult to solve this; initially, ChatGPT might need more time to create a blank blog template, but it will eventually speed up.
Improving ChatGPT's Task Management and Memory
How ChatGPT reads a page, determines completion, and moves to the next task is crucial. Currently, it doesn't review what it writes or retain information about completed tasks, unlike humans who follow a flow and know when to move on. This could be improved by enhancing ChatGPT's memory and task management capabilities.
For example, ChatGPT doesn't know which of the fifteen tasks it hasn't done, nor does it have a memory for job time arrangement. While it can detail fifty pages, it doesn't know how many it has completed or how many more clicks are needed. ChatGPT's unconsciousness in task tracking is a significant limitation.
Human Superiority in Task Management
Humans recognize their work and signatures. ChatGPT lacks this awareness. It doesn't store its writings in memory and doesn't know when it has completed a specific percentage of a goal. Humans know when to move from one task to another and understand the importance of sequence and timing. ChatGPT could benefit from a similar ability to remember completed tasks and manage ongoing tasks effectively.
Future Vision for ChatGPT
Currently, ChatGPT can correctly provide fifty file types for a Ruby on Rails website skeleton, but it can't read all fifty pages as a whole or double-check like a human engineer. Small missing details can prevent the formation of the skeleton structure. Enhancing ChatGPT's conscious memory to comprehend its writings could solve this.
Imagining ChatGPT's capabilities two years from now, its current slow processing in such projects should be seen as normal.
Path and URL Management in Ruby on Rails Projects
ChatGPT plans URLs in Ruby on Rails but doesn't know which folder or file extension to check at a given moment. It understands the folder structure but lacks the timing sense to figure out which page to read and which line to control. It can't match Ruby on Rails URLs to virtual machine or personal computer paths. Initially, a human intervention at certain steps might be necessary to smooth the process.
Conclusion
My goal is to become a traditional Ruby on Rails engineer. I envision a future where ChatGPT can autonomously manage tasks like creating a Ruby on Rails website, including downloading necessary files and managing folder structures. ChatGPT's current limitations in memory and task management need addressing, but the potential for improvement is significant.
I appreciate my education at UoPeople and other online resources, which have provided me with a solid foundation in Computer Science. Despite challenges, I remain committed to my goals and look forward to future advancements in AI and software engineering.
Note: Look at the pizza seller here https://youtu.be/vO7Q7Rx2r-8?si=ScpTK2FDf6Ly-5Jp
Only a person but the tasks are so well ordered, chatgpt on its own is not able to this flow. I want such a task job order is planned for chatgpt in Family Physician office, current technology has this only such well timing and task connection needed. For example chatgpt can take new papers from patient scan them send them to MD screen, summarize past main complaints of the patients from past data. Record the complaints, suggest which labs needed, or not needed (An example application is KAHUN) but should fit a real office. Should be ready for specially a family physician office. This is so much easier than the above ruby on rails-chatgpt project. Because a Family-Physician chatgpt intelligence mostly need domain knowledge ır Family Medicine knowledge. Even if a family-medicine-chatgpt is a breakthrough technology it is CS application to domain medical knowledge.
Anyway, an example task for an family medicine helper robotic chatgpt4o is planned as a summary here by chatgpt itself. http://humantranslator.blogspot.com/2024/05/a-family-physician-helper-robotic.html
..
