Wednesday, January 17, 2024

Hitch in the Wagon

Very early on in my writing career, I published an article in a business journal called Pacific Basin Reports. If memory serves, it was an analysis of mining interests in Southeast Asia.

A few weeks after the article appeared, I was startled to read the very same article, word for word, under someone else’s byline in the prestigious Far Eastern Economic Review. I’d been plagiarized and it wasn’t at all subtle.

I wrote a letter of complaint to the Review and in response got a brief apology and a small check — the fee for the article.

Plagiarism is no doubt as old as publishing, but it is seldom that blatant. In other cases I became aware of over the years, writers lifted sentences or paragraphs from published articles to enhance their own work. This happened far more often with college students or very young journalists than veterans, and we often caught the offenders and disciplined them.

The Internet brought the potential for plagiarism to a whole new level, but also the tools for detecting it. Fast forward to today and we have the spectacle of AI, in the form of ChatGPT, being used by CNET to generate entire articles.

While this isn’t plagiarism — it could perhaps be called bot-ism — it is a danger nonetheless for honest journalists. One more in a long litany of dangers. Technologies have already helped destroy the economic security of millions of jobs in publishing over the past quarter-century.

Now writers face the prospect of being replaced, as do many other workers, by robots. The fact that they can’t do our jobs anywhere near as well as we can may not be relevant. The evidence is mixed as to how much Americans value great writing in the first place. It is a very tough way to make a living.

Meanwhile, we’ve already gotten used to bots messing up our sentences by “correcting” us by inserting errors.

Which reminds me of a story: A priest, a minister and a rabbit go into a bar. When the bartender comes over, he takes the orders from the first two but hesitates at the third member of the group. "Aren't you a rabbit -- what are you doing here?"

"I'm only here because of auto-correct."

(I first published this essay last year in January.)

HEADLINES:

  • Trump is facing a second E. Jean Carroll trial. Here’s where the cases collide. (MSNBC)

  • Trump Unleashes On E. Jean Carroll Ahead Of Trial To Determine Monetary Damages (HuffPost)

  • Democrats seize on Iowa results to campaign on threats posed by Trump (Guardian)

  • Trump’s Iowa win will show how badly he has broken the Republican Party (The Hill)

  • Trump's embrace of far-right activist Laura Loomer worries his allies (NBC)

  • Google layoffs continue with ‘hundreds’ from sales team (The Verge)

  • A warning shot over the last mile in the inflation battle (Financial Times)

  • Supreme Court weighs conservative plea to weaken federal agencies (NBC)

  • Iran Launches Missile Strikes in Iraq and Syria, Citing Terrorist Attacks (NYT)

  • Gaza urgently needs more aid or its desperate population will suffer widespread famine and disease, the heads of three major U.N. agencies warned as local officials said the death toll in the territory topped 24,000. [AP]

  • US targets Houthi anti-ship missiles in new strike on Yemen, officials say (Reuters)

  • Houthis, Undeterred by Strikes, Target More Ships in Red Sea (NYT)

  • Qatar announces deal to allow delivery of medicine to Israeli hostages, humanitarian aid to Gaza (AP)

  • China Deals Major Blow to Russian Economy (Newsweek)

  • In the new Afghanistan, it’s sell your daughter or starve (WP)

  • Cloned rhesus monkey created to speed medical research (BBC)

  • Davos updates: Global leaders discuss AI adoption and potential threats (CNBC)

  • The huge open question for business leaders on AI economics: 1980s or 1990s? (Axios)

  • As robot baristas and AI chefs debut at Vegas tech show, casino union workers fear for their jobs (LAT)

  • Google AI Researchers Introduce DiarizationLM: A Machine Learning Framework to Leverage Large Language Models (LLM) to Post-Process the Outputs from a Speaker Diarization System (MarkTechPost)

  • Runaway bureaucracy could make common uses of AI worse, even mail delivery (The Hill)

  • Iowa Restaurant Patron Can Remember Every Breakfast Ruined By Presidential Candidates (The Onion)

 

No comments: