5 Reasons Why AI Is Outpacing Society's Ability to Adapt

It isn’t fun and games anymore. Here are 5 reasons why:

• Belgium newspapers report a man has committed suicide after having conversations with a ChatGPT-like chatbot for an extended period of time. His wife, who spoke with journalists from La Libre, said she was worried about her husband's mental state before he began his intense conversations with the chatbot, but is convinced that he would still be here if it hadn't been for these exchanges.
  
  The story is part of a broader trend, like the one about Replika’s AI companions leaving users heartbroken after updates had been made to the systems without prior notice. The suicide case would be the first reported case of ‘death by chatbot’, a prediction first made by cognitive scientist Gary Marcus in december 2022.
  

• More than 1,000 industry experts, researchers and engineers worldwide have joined a call for an immediate pause on the training of large scale AIs for at least six months, so the capabilities and dangers of systems such as GPT-4 can be studied and mitigated.
  
  This pause should be used to “jointly develop and implement a set of shared safety protocols for advanced AI design and development that are rigorously audited and overseen by independent outside experts.” The initiators make it explicit this does not mean a pause on AI development in general, merely a stepping back from the dangerous race to ever-larger black-box models with emergent capabilities.
  

• Stanford researchers were able to use the recently leaked LLaMA 7B model combined with GPT-3 to create a fine-tuned model of their own for roughly US$600 dollar which was coined Alpaca 7B.

  
  Albeit impressive, it isn't hard to fathom the implications of this feat, since the process is easily replicable for anyone with sufficient machine learning knowledge. It basically means that a potential unlimited number of sufficiently powerful, uncontrolled language models can now be set up at marginal costs.
  

• On Monday, the FTC published a report on voice cloning scams, warning people that their desperate friend or relative on the other end of the phone asking for money may actually be an AI simulacrum of their voice wielded by a scammer.
  
  With as little as a minute’s worth of audio of someone’s voice, AI can now generate a clone of that voice convincing enough so that it can be used to target the elderly, loved ones, relatives or family members. It takes little to no imagination to extrapolate that into a future where digital voices combined with the power of advanced language generation systems will be used by bad actors for much more than cheap scams.
  

• Text-to-image models like the latest version of Midjourney are now able to produce photo realistic outputs of real people. Some examples of have gone viral last week include Donald Trump being arrested, the Pope in a puffer jacket, and Obama and Merkel holding hands.
  
  On first glance, this might come across has harmless, funny and slightly absurd imagery. However, one could ask themselves how easy it is, with the intent to do harm, to weaponize this technology to destroy reputations and spread false falsehoods about celebrities, world leaders and the like.

I repeat: it isn’t fun and games anymore. AI is changing the world faster than we can handle and we, as a society, are not ready.

---

---

Jurgen Gravestein is a writer, business consultant, and conversation designer. Roughly 4 years ago, he stumbled into the world of chatbots and voice assistants. He was employee no. 1 at Conversation Design Institute and now works for the strategy and delivery branch CDI Services helping companies drive business value with conversational AI.

Reach out if you’d like him as a guest on your panel or podcast.

---

Appreciate the content? Leave a like or share it with a friend.