5 Tips to Building a Winning Digital Remote Assessment Centre Experience

ChatGPT is more extroverted, conscientious, and more open to experiences than the average working age human, new research from Sapia.ai, the world’s only smart chat platform powered by AI, has found. 

If these generative AI models were job candidates responding to interview questions, what kind of personality would they project in their natural form? ChatGPT (based on GPT-3.5) and more recently GPT-4 are built using a process that includes “reinforcement learning from human feedback (RLHF)” to produce responses which are less likely to make up facts, be toxic or harmful in their sentiment compared to the earlier GPT-2. Could this friendliness and politeness be verified through the personality these models project? These were some of the questions Sapia Labs team set out to answer. 

“We all know ChatGPT can be prompted to respond in different ways and, as AI models, personality is not directly applicable to them. However, given these models generate responses similar to humans, we wanted to better understand the personality projected by these models when they are not prompted to be a certain personality, or in its natural form, and be able to distinguish its responses from that of a human,” Sapia.ai’s Chief Data Scientist Dr. Buddhi Jayatilleke said.

The study, a first of its kind, analysed over 6,000 responses from GPT-2, ChatGPT (GPT-3.5) and GPT-4, and cross-examined them with Sapia.ai’s dataset of over 2.5 million candidates across 47 countries with over 1 billion words shared by job candidates. Sapia Labs used their industry-leading personality inference models based on InterviewBERT, a fine tuned version of Google’s BERT large language model to infer the personality dimensions similar to the well known HEXACO model of personality. 

Their findings provide fascinating insights into the nature of these generative models. Both ChatGPT and GPT-4 scored significantly higher on the dimensions of honesty/humility, agreeableness, and consciousness compared to GPT-2. These results align well with OpenAI’s description of how ChatGPT models differ from earlier GPT models. The higher honesty/humility and agreeableness is consistent with modesty, politeness, and friendly responses. Additionally, the greater likelihood to follow directions and provide accurate information aligns with ChatGPT’s higher conscientiousness. But the most interesting finding is that both ChatGPT and GPT-4 showed significantly higher levels of extroversion, conscientiousness, and open to experience compared to the human benchmark dataset. In other words the newer GPT models trained with human-in-the-loop project a more sociable, open minded and diligent personality.

Global brands including Woolworths Group, Qantas and Spark NZ trust Sapia.ai to accelerate and enhance their recruitment and promotion processes. A conversational, Natural Language Processing (NLP) based AI chat interview, assesses and screens for the best talent at scale via an easy to use messaging platform.

In addition to improving diversity outcomes by eliminating unconscious bias, it also allows companies to re-allocate thousands of hours spent screening talent towards higher value tasks. 

Hello HR Tech – Recruiting is so much easier!

Last week I made a promise to share a journey that brought me to be working in the business at the cutting edge of technology and science within the People/Talent sector.

In my previous post, I shared some of the thinking of people within my sector. This is what I learned about hard work during my 13 years working in tech recruitment.

I was 22 years old when I became a recruiter. I was competitive, driven and hungry to succeed. Not only in financial terms, like many other recruiters, but also my professional status and standing. I wanted to be one of the best at my job and to be respected for the work I did.

And I know there are thousands of recruiters out there whose hard work often goes unrecognised by clients, candidates, managers and colleagues alike. I no longer know exactly what it’s like to be a recruiter in 2018 but back in 2005-2010 if you joined one, my teams, we’d have had conversations that went something like this:

There is no shortcut in recruitment

It requires a lot of hard work and skill with a splash of good luck.

The hard work is the time commitment needed to consistently deliver for your clients and candidates.

You need the skill to learn the difference between C# and C++ and how technologies stack together.

Eventually, your business development efforts will combine with good luck when that client answers your call and confirms they are indeed looking to hire someone within your vertical specialism. Happy days!!

You agree to terms for the customer’s key role, you pat yourself on the back and then you go again – back to the hard work because now you’ve got to find suitable candidates.

Good recruiters already have a network of great candidates – you go to them first, qualify/rule out and you’ve got a shortlist inside an hour or two. Then, more hard work.

When the other unknown recruiters working at unknown agencies also trying to fill the same role, clock off at 6 pm to enjoy their evening plans, you’re still in the office.

If you’re anything like I was you’ll still be in the office until 9 pm when the contractors start to get a little irate.

“Sorry for ringing so late in your evening but I’m trying to fill a key role for an important customer.”

Most of them appreciate your hard work and candour. Some even sound impressed with your commitment.

A few get grumpy but them’s the rubs – it’s water off a duck’s back for a driven, professional recruiter who wants to do their best for their customer and won’t mind, professionally, ruffling the feathers of a few early-to-beders to ensure they keep on top of their game, delivering great candidates to their clients.

Eventually, your hard work pays off and you place the successful candidate (probably after at least one candidate did an interview no-show following the death of a distant relative/hospital appointment/dog vs homework / insert obscure excuse)

Meet Tom & Sally to get a sense of what I was filling – I was definitely ‘Tom’! 

That was my early recruitment career. Because I knew there were no shortcuts to success. I needed to graft, sacrifice my evening socialising (don’t worry, I made up for it at the weekends!) to ensure I found the best candidates for my clients.

I was a recruiter and I really, really loved my job. I genuinely hope today’s recruiters love their jobs as much as I did but the recruitment world I knew is no longer. And that’s because Talent AI has created a shortcut!

Not a corner-cutting shortcut. But an evidence-based, efficiency-creating, quicker, faster, more accurate shortcut.

AI can now rapidly identify suitable talent and create a shortlist of candidates for a human recruiter to then engage with.

A shortcut that also helps remove bias from talent workflows.

In fact, it’s such a clever shortcut that it should have its own name. I have a suggestion. Let’s call it…Recruitment!

Because recruitment was still recruitment when ATS providers rolled out filters and keyword identification tools which were quickly gamed by candidates – writing retail on a CV pushed it up the results list but that didn’t make the candidate more knowledgeable in retail.

It’s still recruitment – just done heaps better

Recruitment was still recruitment when talent attraction projects were created. Recruitment is still recruitment throughout the modern-day careers day (which I hope has evolved from my experiences back in the early 2000s)!

It’s still recruitment if you bring in video interviews (disclaimer: I hate the idea of video interviews; I think they simply shift bias to a different stage in the recruitment process).

Recruitment will still be recruitment with AI, it’ll just be better for candidates, clients and recruiters alike.

Suggested reading:

https://sapia.ai/7-tips-to-making…stment-decisions/

PredictiveHire – In the press

An Australian tech firm that uses artificial intelligence to quickly filter job applicants has managed to successfully complete a Series B fundraising round, despite financial markets catching coronavirus last month.

Melbourne-based PredictiveHire finalised the $3 million raise, led by recruiter Hudson with returning venture investors Rampersand and Capital Zed, just after bans on mass gatherings were announced and segments of the economy began shutting down.

“Some of our FMCG [fast-moving consumer goods] retailers, contact centres and emergency services clients have been getting thousands of applications per vacancy in the past couple of weeks, and they need a tool that can filter them quickly but with humanity,” Ms Hyman said.

PredictiveHire claims to have nine of the ASX 100 among its clients, including Wesfarmers-owned Bunnings, who use its software-as-a-service to text questionnaires to the mobile phones of job applicants.

Encouraging 50 to 100-word answers to a handful of questions like ‘what is a change in your life that has happened to you and how did you deal with that change?’, PredictiveHire runs each submission through an engine that performs a branch of artificial intelligence called natural language processing. Built on 25 million words texted back by 350,000 applicants for previous jobs, PredictiveHire claims its engine – run by co-founder Buddhi Jayatilleke, who built the data science team of human resources tech unicorn CultureAmp – can automatically provide hirers with the applicants that best suit their pre-set criteria.

Avoiding Bias

“There’s a lot of ways people can game CVs, but it’s the words and responses to relevant questions that give a real insight into a candidate’s suitability,” Ms Hyman said. She admitted there was little the start-up could do about applicants who get someone else to answer the questions for them but relied on that being picked up by the phone calls or face-to-face group interviews that followed on from PredictiveHire providing its shortlist.

“It’s designed so that your best chance of success is being yourself,” Ms Hyman claimed. If English is your second language, there’s no need to worry because we’re not biased against that, or race or gender or address or any of those factors that work against diversity when hirers take the CV-reading approach,” she said. For the thousands of applicants that will inevitably be unsuccessful as the COVID- 19 crisis raises unemployment, PredictiveHire provides automated feedback including six insights into their personality and a coaching tip for future interviews.

“Even in a usual year, the big hirers reject in six figures, and these people are also their customers,” Ms Hyman said. “They want to give them a good experience and constructive feedback, but there’s no way that’s going to be done consistently for every candidate using manual processes.”

PredictiveHire will use the $3 million injection, which takes its total raised to $5 million since launching two years ago, to further its push into graduate recruiting. This more sophisticated process, only possible as its proprietary data bank of words had grown, was still in demand even as the pandemic stalled markets, according to Ms Hyman. “Good employers can see to the end of this and still want the best talent as it becomes available,” she said.

Suggested Reading:

https://sapia.ai/blog/the-impact-of-picking-the-wrong-assessment-is-measurable-and-high/