Photo Credit: pexels
In June, the San Francisco-based artificial intelligence research lab OpenAI officially released a commercial version of its GPT-3 language generation model for private beta testing, and things have taken off from there. Almost everyone agrees the language generation tool is the best the world has ever seen, and it’s already near impossible to distinguish text generated from this tool from text generated by a human. In fact, a bot using GPT-3 was able to hang out on Reddit for a week, posting lengthy and comprehensive answers to user queries, with only a few AI researchers able to figure out what was really going on. So it’s only natural to ask: How will GPT-3 change blogging and social media?
Bots will become the new content creators
First and most obviously, it’s clear that bots will become the primary content creators on the web – and that includes social media platforms as well. Already, bot content dominates platforms like Twitter, and it’s easy to see how GPT-3 will enable bots to take over other social media platforms as well. What is so powerful about GPT-3 is the fact that it is a general-purpose “text in/text out” model, meaning that it can be used for any English-language task you could imagine. Already, people have used GPT-3 to write poems, respond to user queries on Reddit, and generate some incredibly sophisticated prose text on weighty subjects such as racism, suicide, conspiracy theories, and immigration.
GPT-3 is not a “one use case” model for a specific niche or industry, and that opens the door for GPT-3 to engage in tasks that have previously been beyond the reach of bots. Moreover, add in the fact that it’s a lot cheaper to pay a bot (where the only cost is the cost of licensing the technology from an entity like OpenAI) than a team of content creators, and it’s easy to see how human content creators – editors, writers, bloggers – are destined to become historical anachronisms,
Microsoft could become the next big social media player
OK, OK, we’ve been hearing about this for years, but Microsoft has been looking for ways to become a major player in social media. The latest idea – floated just this year – was that Microsoft would somehow acquire Chinese social media giant TikTok and use that acquisition to become the hip new social media platform for young 20-somethings. Well, that didn’t work out, so guess what? Microsoft decided to play around with GPT-3.
In September 2020, just months after OpenAI announced the commercial availability of GPT-3, Microsoft signed up for an “exclusive” license to use GPT-3. So here’s the big idea: Microsoft will slowly begin to include GPT-3 functionality in all of its products. Imagine being able to chat with recruitment bots on LinkedIn, for example. And Microsoft has already said that it’s cutting its human editorial staff at its MSN news service, as part of a broader initiative to embrace AI in content creation. So it’s a no-brainer to see how GPT-3 fits into that overall narrative. Who needs humans writing about sports, entertainment and politics when you can have some GPT-3 bots doing so?
The next big social media job title will require AI knowledge
Currently, if you want a job in social media, you don’t need any knowledge of AI, deep learning or natural language processing (NLP). All that could change if GPT-3 takes off. Instead of hiring college grads with a cool Instagram site and high-end blogging skills, recruiters will be looking to hire college grads with some experience in deep learning and AI. After all, they will need people who can “train” the bots and get them to respond in the right way. While GPT-3 offers a very flexible “text in/text out” interface, you still need to have some sort of oversight of what is being produced. And this opens the door for all kinds of “next generation” services based on what’s already available today. Take WebMD, for example. Lots of people visit sites like these to get informal medical and health advice. Now imagine being able to chat directly with a “doctor GPT-3 bot” able to speak to your specific ailment or complaint. (Just be sure to get a second opinion from a human!!!) On Reddit, the GPT-3 bot was already to provide psychological counseling about suicide, so nothing is off-limits here.
Overall, it’s an exciting time for AI and its possible applications for social media and blogging. The bots are here and they are here to stay. The quality of text is already extremely high, and it’s natural to assume that it will continue to get higher with each passing year. In just a few years’ time, we could be facing the prospect of an Internet completely dominated by these text-producing bots.