The hype surrounding ChatGPT is surely at its prime. And with good reason. The sudden jump in capability is almost hard to comprehend, and a wealth of opportunity has opened up. The race is now on, both in terms of developing even better language models, but also on building valuable applications on top of this technology.
So, what is this ChatGPT thing and what can it do?
ChatGPT is one of several large language models (LLMs) trained to engage users in human dialog by predicting the next word in a sentence with such accuracy that it is sometimes hard to separate from a response from a human. It has been trained on huge volumes of data which has led to its generalizability and added ability to solve tasks that it was not specifically intended to do.
The model holds a lot of handy tricks up its sleeve and can accomplish most text-related tasks you throw at it with decent results. For recruiters, this means, for example, generating drafts of job ads, candidate communication messages, outreach emails, generating relevant interview questions, summarizing resumes, and many other tasks that are needed on any given day of recruiting.
For candidates, it also opens interesting options. For example, generating instant resumes with a good chance of being shortlisted or passing a coding test for a $183k per year salary job for you.
The first iteration of most texts is often a bit bland, but by using your imagination to further hone your request, the outcome can become surprisingly good. Following up with additional prompts can deliver the exact right kind of vibe. You can even ask it to make sure a job ad copy speaks to all candidates in an inclusive way.
Just make sure you double-check everything first..
And in case you’d like to catch the latest trend of announcing mass layoffs via email memo, examples we’ve seen recently from Google, Facebook, Amazon, and others, look suspiciously close to what’s been produced from ChatGPT.
But its capabilities don’t stop at generating text. It can also analyze and simplify information for you.
For example, if you have collected lots of open-ended responses from an employee feedback survey, you can use ChatGPT to identify the most recurring feedback in a much better way than most survey tools can.
Other use cases include devising business plans, writing poetry, generating ideas, and providing sentiment analysis for example.
For one thing, ChatGPT is a huge step in the usefulness of conversational agents and is likely to unleash a whole new generation of powerful tools that will help achieve tasks more efficiently.
It also means that many of the tools you use today will soon be outdated. And switching to a more efficient option powered by an LLM can bring a huge difference. According to the prophecy by Sam Altman (CEO of OpenAI), the companies that are most likely to be successful end-users of LLMs are those in the ‘middle-layer’ with a few common traits.
- Lots of well-defined and sector-specific data
- Ability to train their own models using specific data on top of the LLM framework
- Great tech infrastructure to be able to incorporate new tech easily.
They won’t have to have created the base model but they will have created something they can use just for themselves or share with others that has this unique data flywheel going that sort of improves over time and all of that. So I think there will be a lot of value created in that middle layer.
- Sam Altman - CEO, OpenAI
In other words, the ‘middle layer’-type companies will be able to improve the performance of a large language model in a very specific field.
Some tools are extremely well aligned with this generational shift and have already access to lots of data that a new, domain-specific model can be trained on, on top of an LLM. This will lead to expert knowledge in niche domains, for example, recruiting for high-volume positions, labor law, compensation questions, and other applications within the HR domain.
Make no mistake about it. Large language models and generative AI like ChatGPT and its future iterations will have far-reaching effects on society, employment, and human job descriptions. The way we work will most likely change drastically as more and more sophisticated AI models pop up and new use cases keep appearing.
Employers are already starting to realize that, even if they can’t replace human personnel altogether just yet, they sure can expect a significantly higher output for professionals that can be helped by generative AI such as journalists, copywriters, and other text-intensive professions.
For example, Buzzfeed’s stock price skyrocketed when they announced they will use ChatGPT to produce quizzes and other light-weight content in the imminent future to drive profitability and decrease the number of human workers.
Humans that can leverage ChatGPT to be more efficient have lots to win. In fact, cooperating successfully with these kinds of models could become a necessity rather than a sought-after skill faster than expected.
Want to stay in the know about Conversational AI, LLMs and how it all will impact the recruiting sector? Subscribe to our newsletter.
