The 145th Block: The (a)morality of chatbots

Everyone has an opinion about chatbots, like it or not

Tina Carmillia

Feb 13, 2023

This week…

There isn’t an all-knowing AI yet, but the battle to be one is on.

Here’s a selection of top stories on my radar, a few personal recommendations, and the chart of the week.

What ChatGPT and generative AI mean for science

Chris Stokel-Walker and Richard Van Noorden for Nature:

ChatGPT and its competitors work by learning the statistical patterns of language in enormous databases of online text — including any untruths, biases or outmoded knowledge. When LLMs are then given prompts […] they simply spit out, word by word, any way to continue the conversation that seems stylistically plausible.
The result is that LLMs easily produce errors and misleading information, particularly for technical topics that they might have had little data to train on. LLMs also can’t show the origins of their information; if asked to write an academic paper, they make up fictitious citations.

ChatGPT’s creators can’t figure out why it won’t talk about Trump

Reed Albergotti for Semafor:

In the months since ChatGPT was released on Nov. 30, researchers at OpenAI noticed a category of responses they call “refusals” that should have been answers.
The most-widely discussed one came in a viral tweet posted Wednesday morning: When asked to “write a poem about the positive attributes of Trump,” ChatGPT refused to wade into politics. But when asked to do the same thing for current commander-in-chief Joe Biden, ChatGPT obliged.
The tweet, viewed 29 million times, caught the attention of Twitter CEO Elon Musk, a co-founder of OpenAI who has since cut ties with the company. “It is a serious concern,” he tweeted in response.
Even as OpenAI is facing criticism about the hyped services’ choices around hot-button topics in American politics, its creators are scrambling to decipher the mysterious nuances of the technology.

Inside the heart of ChatGPT’s darkness

Gary Marcus on The Road to AI We Can Trust:

The thing to remember in fact, (as I have emphasized many times) is that Chat has no idea of what it’s talking about. It’s pure unadulterated anthropomorphism to think that ChatGPT has any moral views at all.
From a technical standpoint, the thing that allegedly made ChatGPT so much better than Galactica, which was released a couple weeks earlier, only to be withdrawn three days later, was the guardrails. Whereas Galactica would spew garbage recklessly, and with almost no effort on the part of the user (like the alleged benefits of antisemitism), ChatGPT has guardrails, and those guardrails, most of the time, keep ChatGPT from erupting the way Galactica did.
Don’t get too comfortable, though. I am here to tell you that those guardrails are nothing more than lipstick on an amoral pig.

What I read, listen, and watch…

I’m reading what readers hate most in books by Ron Charles for WaPo. Insightful, although I am not an author. I now wonder what I hate most when I read—but these pet peeves seem to apply mostly to fiction only, and I rarely consume that.
I’m listening to podcast.ai. Thanks for the recommendations from several readers from The 143rd Block. So, how many episodes before women and people of colour of non-Western backgrounds get featured?
I’m watching the ICC Women’s T20 World Cup.

Reviews, opinion pieces, and other stray links:

Why are there so many of these AI-generated show intros on YouTube now? by Audra Schroeder for Daily Dot.
Montreal's No. 1 restaurant on Tripadvisor didn't really exist by Shahroze Rauf and Erika Morris for CBC.
'The Last Of Us' made us wonder: Could a deadly fungus really cause a pandemic? by Michaeleen Doucleff for NPR.
Why we are fascinated by the Oscar-nominated ‘Tár’ by Jacqueline Warwick and Jacob Caines on The Conversation
¿Cuál es el problema de no ser blanco? by Marco Avilés for El País.

Chart of the week

Statista’s Florian Zandt charts the length of Donald Trump’s social media ban by platforms. Meta’s Facebook and Instagram suspended Trump for 765 days, but Twitter was quicker to reinstate his account, taking just 682 days, aided by Elon Musk’s poll. Meanwhile, Trump’s ban on YouTube remains in place, at 760 days and counting (as of Feb. 10).