By Melissa Towns, Project Manager / Virtual Assistant at Virtual Assist USA
AI (Artificial Intelligence) is the buzz in the tech world these days. AI is human intelligence that is processed by computer systems that function by absorbing large amounts of data. The data is analyzed to find correlations and patterns and then uses that information to make estimates about future states. For example, a chatbot can be fed samples of information to learn lifelike exchanges among humans. Read on for our thoughts about the future of Virtual Assistant work.
Although AI has the potential to positively impact how human Virtual Assistants work, AI tools have their downfalls. While much of the work that human Virtual Assistants do is tech-focused, it is also collaborative. And that is something AI cannot replicate. ChatGPT, created by OpenAI, has shown to be well-informed when there are excellent data to learn from, but it cannot completely replace the personal touch of humans. OpenAI’s Chief Executive Sam Altman tweeted: "We have lots of work to do on robustness and truthfulness."
The catch with ChatGPT is that it essentially doesn’t know anything. It’s only trained to discern patterns that it’s gathered from the internet. Therefore some of the information it gathers could be incorrect. In certain situations, ChatGPT will incorrectly interpret the question and be unable to ask for more context.
A real-world example we’ve recently encountered is websites such as StackExchange, whose purpose is to provide correct answers to everyday questions, which have unfortunately been flooded with incorrect information generated by ChatGPT. Human volunteers were tasked with sorting through the responses generated by ChatGPT, so they could maintain high levels of accuracy. To prevent further damage to StackExchange's reputation, all answers generated by ChatGPT have been permanently banned from its website.
While ChatGPT and other AI tools are fun to use, if you believe that AI as it exists today can replace human content writers – well, read this.
AI may be easy and exciting to use, but remember that a human Virtual Assistant brings an edge and personalized touch to your business that AI cannot. It’s vital to review and make sure whatever information is generated is factually correct. Spreading false information has consequences, and although it’s uncertain how exactly ChatGPT is fed its data, humans will have no other choice than to blindly trust that OpenAI will continue to develop ChatGPT ethically.
Another example: Google Voice may have cut errors in transcription by almost half in 2017, but the human Virtual Assistant remains unmatched when it comes to language processing and decoding voice instructions. AI is progressing at a rapid pace but still falls flat at the element of troubleshooting– whereas human Virtual Assistants can handle any project, event, or glitch, such as cancellation or delay, without missing a beat.
Human virtual assistants can function - and work - independently. AI assistants cannot at this point – as they lack individuality, innovative thought processes, and decision-making capacities. There can be a collaboration between your human Virtual Assistant and AI in the near future– and you will likely save time and money by doing this. For example, you can instruct your VA to reduce time spent on a project by using AI to accomplish research. But when it comes to strategy, innovation, forecasting, and idea generation – a human VA can’t be matched.
Such work done by your Virtual Assistant is an incomplete science -- here, 1+1 doesn't always = 2. There are a lot of soft skills involved, as well as troubleshooting and variations. These variances simply cannot be fixed into a computerized system and replace the human brain.
References:
Comentarios