IT Industry journey from information processor, to deploying IT workforce for business & developing limited intelligent version of human!

Summary Note: Automation-> Artificial Intelligence-> Artificial Gen AI-> Trained AI Model to do all human works in office, business of all areas->Refined, trained and better Humans!

IT Industry has registered growth in "deep data & code analysis" science of self learning & decision making in leaps and bounds in the age of evolution of processors, memories, and converge infrastructure development. This led to finally new style of computation which also appears like complex system data processing, shinning and unearthed in typical Bollywood style to world in shape of "artificial intellisense" to "artificial intelligence".

Further course of intelligence took shape of refined form of "Artificial GEN AI".

Just like every events are different so solution to it are different. IT Industry core solution model still helps all of us and paved way to close that gap while providing solutions, as for every events, situations are different so custom solutions will govern all solutions requirements. This can be validated with same size solution can't be assigned to all business requests with different requirements.

This custom solution is just keeping all IT professional with hope of keeping job secure, as this has option to give job to all IT professional or all jobs would have been executed by robots by putting one script to do same job for all requirements.

Let's see in summary how these phases went in front of our eyes.

Automation

Business, efficiency, intra company competition, on-going evolution of processors, memory and shrinking cum optimization of data transfer requirements pushed intensely IT industry or IT enabled organization to review, and implement solutions based on automation.

Pushed from leadership on automation, didn't only bring "automation" led solution in organizations, however there were constant and honest attempts, made to do optimization in business and technology and that inspiration cum blessing became new norm in IT Industry. This led to bringing all newly discoveries based processors, memories, servers and converge infrastructure started played critical role in spinning up of many revised version of automation solution. These many resources of automation means, led developers and system admins to start writing automated solution with intensity to unearth next version of automated solutions.

This new era of automation solutions were too different from traditional mode of automations to self educating automated solutions as well. This means all new solutions started to have code piece to store learnings based data into database or system or script code to use new learning into next iteration of code execution flow and allow decisions to be taken inside code while code execution taking place, rather than seeking developer, system admins to intervene and allow next iteration of code execution work-flow to receive instruction, and finally then let code work-flow to move in certain expected direction. This was intelligence which was already brought into automation phase of IT Industry.

This stage or phase's state of automation helped to halt IT industry to decide further course of action based on how to decide to allow code to learn from on-going code executions in terms of learnt events along with direction in midst of code execution. This halt was required and felt by IT Industry leaders as they received input from all automation champions that we were in phase to start adding intelligence into automation script which will remove one admin cost, to enable code execution to decide or let code decide on its own while code execution were taking place. This was felt also due to reason as some scripts were required to execute at mid night or late night where system admins were not present in office to provide input to script to complete one cycle of code execution or status of code execution was found as failed in system log.

Artificial Intelligence

"Artificial Intelligence" took birth to address "Automation" phase based challenges and to provide direction to IT Industry as pace of evolution cum refinement at hardware layer of processors, memory and infrastructure convergence were still progressing at great speed which had already taken IT Industry by surprise by giving by-products in terms of "Cloud" mode of operations to solve many issues related to best practices and resources wastage to name few. "Cloud" mode of operations don't fix only such issues, however also came up with solution to address, public, private, services and community based computing requirements from small level organization to highly custom and big size organization.

"Cloud" mode of operations also brought services further segregation for Infrastructure as a service(IaaS), Platform as a service(PaaS), Software as a service(SaaS), Database as a service(DBaaS) and many we wanted and could name it.

Artificial Intelligence phase kept progressing after taking baton from "Automation" phase of IT Industry, in terms of design, developing, refining means of code of self teaching and enabling code execution to decide course of action for "code work flow" inside code and during code execution without waiting for developers of system admins to guide further. This was brought into system on multiple platforms and means at code layer either by storing learnt lessons in database to refer in future execution of code phase or updating code portion to decide course of action in updated version of model(e.g. algorithm teaching or refinement, also called as model training). The storing of decisions or directions to let model move to next phase of code execution differs on technology, platform as well as business requirements. This led to do baseline, identification and establishment of all supporting services, tooling's and platforms to support goal of "Artificial Intelligence"'s deliverance.

So overall below steps are performed on average to develop an AI model: -

1. Define expected outcome from AI model.

2. Gather, clean and validate the data.

3. Create matching or suitable algorithm to match business requirements.

4. Train the model by training the algorithm with each iteration. This is most critical and important part of AI model development.

1. Prepare and refine data.

2. Provide data set in specific pattern to teach model to produce outcome in some specific goal in mind.

3. Bringing iterative improvements in data and model prediction capability with model testing.

4. Finally deploy the newly trained model on required endpoints for target audience to consume its services on agreed delivery business goals, once it produces output as it was expected from business.

This doesn't end on this phase, however it is starting point, from this phase onwards, you could keep refining the model and there will be requirement to deploy the model on target endpoints on network (e.g. standardization of development, deployment environment needs to be in place too).

Note:

1. On tech-stack, data can be in excel or media from where your algorithm could read from. Algorithm could be in language you and your client agreed on, infrastructure, platform and deployment services\models and approach should be aligned in too.

2. Solution will drive which steps to follow, if database based AI model needs to be developed and deployed then excel file based or external file based data once imported into database, same will not be required now and then, untill new data needs to be in system.

3. Database are coming up with required procedurs, objects to develop and deploy AI models.

4. Basis solution requirements, above mentioned steps may vary as per business requirements.

5. Above steps are performed across platforms of cloud or datacenter on average.

Further refinement took place in technique of solution model creation, development and refinement by standardization of Infrastructure, platform, language, data source\destination, solution finalization and deployment on internet or intranet or expected endpoints of all to access and consume.

Artificial Intelligence also brought everyone attention towards below critical aspects: -

1. Business use case & requirement understanding to deliver expected shape, size custom solution.

2. Solution accessibility requirement and hosting, tech-stack requirements in terms of network and security, storage, data security in rest\transit, identity\authorization, solution endpoints, publicly\private mode of access, data integrity cum identity, environment and data pattern to educate solution model and placement of solution model for target audience to consume.

3. Solution development environment and target data pattern consumption to train solution model to predict or forecast or do delivery as required. Possibly usage and enablement of solution development and further platform usage of CI\CD pipelines to achieve final goal of required business delivery.

4. Training and refining solution model to comply with location specific organization and government rules and policies.

5. Refining solution to comply with reports content by applying filtering on abusive and adulterate controls.

6. Final solution deployment to target endpoint.

7. Further plan for keep training solution model and placing on agreed endpoint as each model needed to learn new and custom requirements. This needed entire cycle of model development, refinement to go once again against new changes in expectations.

As IT Industry observed, there had been consistent evolution in areas of processors, memories and infrastructure convergence, there was constant push for refining "Artificial Intelligence" solution model and approach on all layers further. This led "Artificial Intelligence" to break border of prediction\forecast based solution model, to new area of developing content for software code, image development, and many more..

Artificial Gen AI

"Artificial Gen AI", left "Artificial Intelligence" limitation to deploying model of predicting\forecasting solution model to do next level of self-training and delivering content way beyond of just text to text to images, software codes, translation of code across software developing technologies from one to other, Intellisense in Intelligent ways delivering chat messages or content basis record preferences, responding personal and professional emails, to supporting development of cine and films contents and many more products could be built on it.

Responsible "Artificial Gen AI" product and services provider organization further proposed, refined and made standard to put control to comply with data rules and policies across geographies along with compliance with requirement of content free of bias, threat to security, privacy and proprietary property, unexpected content, abusive and adulterate content.

Beyond Artificial Gen AI

Did such above achievements mark the end of AI?, then we forget to limit human mind which is only stopped when he\she was felt to stop otherwise he\she could bring something which always comes with a shocking reality for world to face and handle.

Artificial Intelligence models beyond self-training capabilities

Next possible areas where Artificial Gen AI to move, is to remove all requirements of humans to give them direction to decide next turn of software code execution to take turn, to develop any product from text to do all professional office clerical jobs, develop software on its own, still photo data processing to deploying end to end film\movie, doing business to disrupt entire world and push new ways of business across world, AND gradually move towards deploying human, refined humans and even better intellisense based intelligent models to outclass humans on all required parameters on all areas of human life.

But still will future models be able to match human emotions or they surpass, is something needs to wait for future to tell us. World market is run and driven by human emotions, so this will be a good and relevent angle to watch. So, there will be question, does human understand its own emotions completely?, I guess then only he\she could train any model, as models are made out of huge investment in in-memory, processors, and fast data-based bandwidth converge infrastructure. As managing all such decisions by model when data and code execution takes place will require huge memory, powerful processors and converge infrastructure to function in expected manner!

Hope possibly then limitation of such model will help appreciate & identify importance of God one day! I guess understanding and controlling human emotions will possibly start all eyes open and will there be need of further refinement will be something we will check then!

Science and Spiritualism works on different dimensions, however if there is any clash between these two realms then that will be interesting to experience.