Put 100 recruiters in a room and ask them their least favourite part of the job. I’m willing to bet the vast majority would say writing job adverts or CV coversheets. For those that are active with their personal branding, they may say thinking of, and writing content.
Many soldier on themselves, with others outsourcing these tedious parts of the job with varying results.
So, a technology that could do all this for you in a matter of seconds, should be on a recruiters’ radar, right?!
What is GPT-4?
GPT stands for Generative Pre-trained Transformer, with GPT-4 being the 4th iteration of the technology. In essence, GPT is a neural network machine learning model trained using internet data to generate human like text.
In Lehman’s terms, it’s an AI technology that when given a few starting points, can generate human like text with a wide range of use cases.
“This text-generation tool can create questions, answers, summaries, translations, classifications, code generation, and AI conversations” (https://bloggingtips.com/gpt-4/)
If you’re not familiar with the previous iterations of OpenAI’s language prediction model, here’s a brief history: GPT-1 was developed in 2016. It could predict which word would come next in a sequence. The second iteration of GPT, also known as GPT-2 (which was released in 2018), could make more accurate predictions than its predecessor and even hold a conversation with humans.
The newest version of this model, called GPT-3 (released in 2019), was able to do much more than the previous versions; it could now write short passages from scratch that were described by an expert evaluator as being “very persuasive and realistic."
GPT-4 is due to release sometime soon and has been reported to use up to 100 trillion machine learning parameters compared to GPT-3’s 175 billion, making it significantly more powerful.
Who develops GPT?
GPT is developed by a company called OpenAI. OpenAI is a non-profit organization that was founded in 2015 by Elon Musk, Sam Altman and other tech leaders with the aim of developing artificial intelligence for the benefit of humanity.
It also counts Microsoft as a significant investor.
They work with experts in the field of artificial intelligence to develop new technologies and then release them as open-source models. These technologies have included things like GPT-2 and GPT-3, which were used to create captioning systems that can generate captions for images or videos more quickly than humans can do themselves.
Why should recruiters be excited?
Well for a start, up until this point 90% of this article was created by a GPT-3 engine and took less than a minute to generate.
I’ve previously used GPT-3 engines to help generate ideas for articles and while not perfect, it provides a fantastic starting point to get your creative juices flowing. It can really help with writers block!
The current technology certainly has it’s limitations, and it’s not quite ready to be used as a standalone tool for writing content. However, that may all change when GPT-4 is launched.
As mentioned earlier, GPT-4 is reported to be significantly more advanced with up to 500 times more capability. “To put this into context, GPT-4 will have as many parameters as the brain has synapses” (https://www.hallaminternet.com/what-is-gpt-4/)
Recruitment use cases?
The most obvious and useful one for me is writing of job adverts. It’s a tedious task that many people dislike, and are frankly poorly skilled and/or trained in.
Look at most recruitment job adverts, and they are a copy and paste of the job description with an intro consisting of “our client is a leading…”.
90% of recruiters fail to understand that a job advert and job description are two different things.
Autogenerated GPT-4 adverts could really add a new dimension for recruiters and help them create interesting and compelling adverts.
Maybe it will get advanced enough to write a great CV coversheet in the future?
Another obvious use case is content writing for blogs and articles. I’m really excited to see the advancement of the GPT-4 model as I’ve already had pretty good success with the current version.
GPT-4 could also make searching for candidates easier. It may be possible for GTP-4 to be integrated into ATS search functions to help make predictive Boolean searches and offer alternatives to your search string. Github already has a similar function for software development and uses another OpenAI software called Codex in their AI pair programming function.
Conclusion:
Technology should either enable you to do something you previously couldn't do, or aid you in doing a task faster and better.
Automation and AI technology does exactly that.
GPT-4 has some exciting use cases and I’m sure it will be at the core of some great recruitment tools of the future.
Please feel free to reach out if you would like any tips or advice on autogenerated content!