On Human Value, AI, and Spam

Finally, companies are taking steps to curb AI spam

Ever since tools like Chat-GPT hit the scene, I have been sounding the alarm about the incoming flood of misinformation, spam, and clickbait content. It took a while for platforms and users to catch up, but we’re finally seeing a backlash against low-quality, AI-generated content.

In March, Google announced that it is making changes to combat AI spam in search. They understand the threat to their business. People have noticed that the overall quality of search results has dropped dramatically over the last year. The top-ranked search sites often achieve their search rank by spewing out AI-generated blogs and articles, using back-linking strategies, repurposing expired domains, reputation abuse, and other ways of manipulating search. Of course, this has been possible for a long time, but AI tools have made it scale cheaply and efficiently.

Increasing the Value of Humans

Many industry players jumped on the AI bandwagon and started integrating LLM wrappers into their services and encouraging users to try them. Some sites like Medium, a popular blog aggregator, have pivoted and are cracking down on AI-generated content. Another popular writing site, Fanfare, said, “Any current Fanfare writer that submits an AI-generated story will likewise be shown the exit. No second chances, no redos. That may sound harsh but I want to be completely clear on this.”

More publishers are requiring AI-generated content to be clearly labeled as such. On the consumer side, some encouraging news comes from a recent survey conducted by Mitre that found that “a significant majority of U.S. adults support efforts to enhance AI’s transparency and safety”.

As a content creator and educator, I’ve had learners and peers reach out to me and ask if I’m concerned about AI in my line of work. The answer is an emphatic no! I plan on adding the fact that my learning content is human-generated and curated to my marketing. As people’s trust in platforms that allow AI spam erodes, I predict there will be greater demand for content people create. Even if that content has some flaws, they will be human flaws.

Last year, when Reddit forums went dark in protest, it significantly impacted Google searches. Because of the issues with AI-generated spam, a non-trivial number of searches append “Reddit” to their search query so that they can read human opinions.

Now, Reddit is far from a quality source. I see bad advice and misinformation about technology in the subreddits I follow daily. But like others, I’d rather wade through those responses than be subjected to obvious puff-piece spam blog sites that come to the top of Google search these days.

Where Regulators Must Crack Down

No one should use AI to create and publish content at scale without human editorial processes. How we detect and measure that will be challenging, and it will certainly be an arms race as the tools improve. My biggest concerns are:

1. Political Misinformation: We’ve already seen damage being done by social media platforms allowing bots and misinformation to push political viewpoints to influence elections and stoke discord between citizens.
2. Cyber Security: It is becoming easier than ever to launch phishing at scale, and we’ve already seen some samples of AI being used to create malicious code and automate attacks. Those fake emails and texts are about to get a lot more realistic.
3. Deep Fakes: These should be outright banned, except for those clearly labeled and are satire. Deep fake videos have the potential to be incredibly dangerous.

Using AI Tools Responsibly in Content Creation

Although we’re seeing this backlash start, this doesn’t mean I’ve gone Luddite and don’t use AI tools. Responsible, ethical, and productive ways to use AI in content creation exist. So, I would like to share my take on the dos and don’ts of using AI in your content, whether it’s blogs, videos, or educational courses like those I craft.

Generating Sample Data

AI has helped me by generating sample data for some of my exercises and assessments. For example, when practicing LINQ queries, I had ChatGPT help me generate a list of Luchadores (wrestlers often seen with fancy masks) and their career stats. It did a great job coming up with some fantastical names but not a great job at some stats, like the number of championships (it gave everyone an unreasonable number of championships).

I’ve also used it to replace lorem ipsum content in written samples. This saves time and adds more realism to examples where the main point isn’t the filler content.

Repetitive, Low-Value Tasks

Repetitive and boring tasks that are relatively low value are another candidate for using AI to increase efficiency and your ability to service your customers. Transcription is a big place where I use these tools. For work, I subscribe to Fireflies.ai. Not having to take detailed notes in my meetings means that I can spend all of my time focusing on my client and solving their problems, and it gives me an easy way to summarize and review key points in the meeting later.

Translation

There is huge potential for AI in the translation space. I use it today if I encounter information I want to parse in a language I do not know. I will use it to assist me with converting my human-created content to other languages when the quality reaches an acceptable level.

Processing and Summarizing Data

Another great use of AI for content creators is processing and summarizing data. I want to be aware of trends and happenings on my YouTube channel and social media, but I don’t have time to read everything. Having AI bots monitor trends and bring things to my attention is an excellent efficiency booster.

Promote Your Humanity

There are a few ways that individuals can start to push back against AI spam:

• Add hashtags such as #NoAI or #HumanWritten. Hashtag activism is a thing, and it works!
• Disclose any time AI is used to do anything significant concerning the user’s experience.
• Favor publishing, buying subscriptions, and supporting platforms with sensible AI policies.

If you want help creating great learning experiences by humans, for humans, then reach out!