Introductions
A global trend is emerging where large organisations attribute the rise of Generative AI as an excuse to slow down, halt or remove roles they think are susceptible to replacement by Generative AI. While this shift has sparked concerns about job security in certain industries, innovative and agile organizations instead leverage Generative AI for new business growth.
At AI Singapore, we are users and innovators in AI. Recognizing the potential of Large Language Models (LLM), we are harnessing its power to scale up our signature programs – the 100 Experiments (100E) and the AI Apprenticeship Programme (AIAP). This pioneering application of Generative AI is a game-changer. It scales the impact of our programs and illustrates how Generative AI technology can bring about positive transformations in job creation and talent development.
Increasing The Number Of AIAP Apprentices
Over the last 14 batches, we have seen the following: there is a 50% dropout for each of the stages of the AIAP recruitment. If we receive 200 applications, we will only receive around 50% or 100 submissions for the technical assessment (TA). Of the 100 who submitted, only about 50% will pass. We will then invite these 50 to the interview; typically, only 50% will pass the final interview stage. So, the 25 candidates who are left will receive an offer for the AIAP programme.
So, to grow the AIAP cohort and concurrently maintain the same standard of intake, we can do two things:
- Increase the number of applicants who are not aware of AI Singapore and the AIAP
- Increase the acceptance rate of applicants who apply
Increasing the number of applicants who apply to AIAP is a marketing and outreach activity, and AI Singapore will scale these up.
The AIAP Field Guide And Boot Camp
The AIAP Field Guide is now in its fourth iteration. We have added new content and reorganise the structure for better clarity and ease of use.
From the Preface of the field guide:
This field guide serves several needs:
For aspiring AIAP™ applicants, this field guide provides a structured 12 to 18 months learning pathway to attain the requisite level for acceptance into AIAP™.
For experienced applicants to AIAP™, this field guide provides the necessary knowledge required, and applicants can check themselves against the content outlined here before they apply for AIAP™.
Unsuccessful applicants can also evaluate their skill set against the content listed here and use this Field Guide to deepen their knowledge required before applying to AIAP™ again.
As AI and machine learning are developing rapidly, we will update this content based on our engineer’s experience executing the 100E projects and your feedback
To complement this AIAP Field Guide, the AIAP team have launched the AIAP Preparatory BootCamp.
AIAP Preparatory Bootcamp offers an avenue for aspiring AIAP applicants to review their learnings with a synthetic project and receive mentorship from AISG AI Engineers to accelerate the mastering of the application of their existing knowledge to a project. Our AIAP model is based on mentoring sessions, where AISG AI Engineers lead groups of participants to facilitate sharing and discussion.
All the available spots for the first boot camp in September 2023 have been filled. More runs will be planned.
This article will focus instead on getting more applicants who already know of AIAP and studying to pass the technical assessment.
Ask Kevin! – A Generative AI, LLM-Based Co-Evaluator
Ask Kevin! Ask Kevin! functions as an LLM-based AI Engineering assistant that we will use during the AIAP recruitment and throughout AIAP itself.
AIAP Recruitment Phase
We are developing Ask Kevin! as a co-evaluator to our AI Engineer as a tool to accelerate the evaluation process of AIAP applicants’ technical test scripts during the AIAP application process. This tool helps our AI Engineers assess applicants’ code submissions, assign consistent scores, identify mistakes, and propose recommended solutions. This allows applicants to learn from their mistakes, understand their weaknesses, and improve for the next batch of AIAP recruitment. This innovation in the evaluation process can allow us to handle more AIAP applications as we grow the AIAP program over the next few years.
Deep Skilling And Project Phase
Ask Kevin! Ask Kevin! will be used during the deep skilling and project phase of the AIAP as well. Apprentices using Ask Kevin! can receive immediate feedback without waiting for an AI mentor’s availability, hence accelerating their learning process. This allows AI mentors to oversee larger groups (12-18 apprentices) at a time, up from 4 to 6, increasing our efficiency.
Why Ask Kevin! ? This is because the current Head of AIAP is Kevin and he is the goto guy for any AIAP technical questions.
The Big Board! – A Generative AI, LLM-Based Simulator
Another project the AIAP team is working on is called The Big Board! – which is an LLM-based AIAP Simulator that provides a virtual and immersive learning experience modelled after the AIAP experience.
The simulator will enable apprentices to experience the AIAP in a controlled virtual environment, accelerating their learning while experiencing the development activities and sprints of a real-world 100E project – warts and all. If The Big Board! is successful, it would offer prospective AIAP applicants a better way to prepare for the rigour of the AIAP programme and complete the programme successfully.
This application is called The BIg Board! because our discussion area where we host a lot of discussions, presentations, meals happen is called The Big Board.
Conclusion
Far from reducing our engineering team headcount, Ask Kevin! and The Big Board! will allow our AI Engineering team to grow to support the scale-up of the AI Apprenticeship Programme (AIAP) and 100 Experiments (100E) initiative.
With Generative AI, we will mentor and train more Singaporeans via AIAP and help more companies begin their 100E AI journey!
Ask Kevin! is in Beta, and we will use it for evaluating the AIAP batch 15 recruitment happening this month. The Big Board! The AIAP Team is currently building Ask Kevin! and expects to begin beta testing by the end of this year.
Leave a Comment