A personalized and more persuasive spam will come – El Sol de México

Every day, the messages of princes Nigeriansstreet vendors wonder drugs and promoters of investments safety pins clog customer inboxes emails.

But now, the arms race between the blockers of spam and the senders is about to intensify with the appearance of a new weapon: The artificial intelligence generative.

It may interest you: Do you receive Shein notifications on Instagram? We explain how to block them

With recent advances in AI that ChatGPT made famous, spammers may have new tools to bypass filters, grab people’s attention, and convince them to click, buy, or give away personal information.

As director of the laboratory of RAdvanced Human and Machine zoning at the University of South Florida, I investigate the intersection of the artificial intelligence, natural language processing and human reasoning.

I have studied how IA it can learn individual preferences, beliefs, and personality quirks of people.

This can be used to better understand how to interact with people, help them learn, or provide helpful suggestions. But this also means that you should prepare for a spam smarter who knows your weaknesses and can use them against you.

But what exactly is spam?

Just in case someone doesn’t already know, the spam is defined as emails Unsolicited commercials sent by an unknown entity. The term is sometimes extended to text messages, direct messages on social networks and fake product reviews. The spammers qThey want to nudge you into action: buy something, click phishing links, install malware or changing views.

He spam It is profitable. an explosion of email can generate $1,000 in just a few hours, which costs the spammers just a few dollars, not including initial setup. A campaign of pharmaceutical spam online you could make around $7,000 per day.

Legitimate advertisers also want to push you into action: buy their products, take their surveys, sign up for newsletters, but while a email of a vendor may have a link to the website of an established business and contain an opt-out option in accordance with federal regulations, a email unwanted may not.

Los spammers they also do not have access to the mailing lists to which users have subscribed. Instead, the spammers they use counter-intuitive strategies, such as the “Nigerian prince” scam, in which a Nigerian prince claims he needs your help to unlock an absurd amount of money and promises to reward him very well.

Savvy digital natives immediately dismiss such pleas, but the absurdity of the request can actually select for naivety or advanced age, filtering out those most likely to fall for scams.

However, advances in IA they mean spammers might not have to rely on such unpredictable approaches. AI could allow them to target individuals and make their messages more persuasive based on easily accessible information, such as posts on social networks.

The future of spam

You have probably heard of advances in generative models of large languages ​​such as ChatGPT. The task these generative LLMs perform is deceptively simple: Given a string of text, predict which token comes next. Then predict what token comes after that. And so on, over and over again.

Somehow, training only on that task, when done with enough text in a LLM large enough, it seems to be enough to give these models the ability to perform surprisingly well in many other tasks.

➡️ Subscribe to our Newsletter and receive the most relevant notes in your email

Multiple ways of using technology have already emerged, demonstrating technology’s ability to quickly adapt and learn about people. For example, the LLM they can write entire emails in your writing style, with just a few examples of how you write. And there is the classic example, which is already over a decade old, in which Target she realizes that a customer was pregnant before her.

All new technologies can result in both amazement and danger. The difference lies in who creates and controls the tools and how they are used.

* Assistant Professor of Computer Science and Director of AMHR Lab, University of South Florida.