AI helps, they say. Software makes life easier, they say

First off, tip o’ the hat to Mason for spotting this too.

We are all hearing how Artificial Intelligence, aka AI,  is the wave of the future, and that machine learning/intelligence will lead to breakthroughs, stunning leaps, and making everything better, right? (We all know that in actuality, almost all technology advances are used to kill each other off more easily, right? THEN they think about applying it to normal life.) But AI is the current buzzword, like quantum computing was, like nanotechnology was, like…

There has been more than one fella who has equated Artificial Intelligence with Natural Stupidity. Will one beat out the other? Who knows – but seemingly, some AI seems to have incorporated Artificial Stupidity to blend the two.

Tech behemoth OpenAI has touted its artificial intelligence-powered transcription tool Whisper as having near “human level robustness and accuracy.”

But Whisper has a major flaw: It is prone to making up chunks of text or even entire sentences, according to interviews with more than a dozen software engineers, developers and academic researchers. Those experts said some of the invented text — known in the industry as hallucinations — can include racial commentary, violent rhetoric and even imagined medical treatments.

Experts said that such fabrications are problematic because Whisper is being used in a slew of industries worldwide to translate and transcribe interviews, generate text in popular consumer technologies and create subtitles for videos.

I don’t translate interviews any more, but can swear that subtitles nowadays are pitiful – and sometimes even funny – to the point of becoming next to useless. How will we watch a Scots detective without having his mumblings translated into something approximating intelligible speech?  (“Clarkson’s Farm” excepted, there’s a guy on there that NO ONE understands but his son. Even the subtitler gives up.)

But that could be a bit dangerous in a medical context, especially if “burned herself on a canner” turns into “cancer” or some such. At least it’s rare, though, mmh? Uh, no:

A University of Michigan researcher conducting a study of public meetings, for example, said he found hallucinations in eight out of every 10 audio transcriptions he inspected, before he started trying to improve the model.

A machine learning engineer said he initially discovered hallucinations in about half of the over 100 hours of Whisper transcriptions he analyzed. A third developer said he found hallucinations in nearly every one of the 26,000 transcripts he created with Whisper.

“Nobody wants a misdiagnosis,” said Nelson, a professor at the Institute for Advanced Study in Princeton, New Jersey. “There should be a higher bar.”

Whisper also is used to create closed captioning for the Deaf and hard of hearing — a population at particular risk for faulty transcriptions. That’s because the Deaf and hard of hearing have no way of identifying fabrications “hidden amongst all this other text,” said Christian Vogler, who is deaf and directs Gallaudet University’s Technology Access Program.

Think this doesn’t affect you? Ever hear of ChatGTP?

While most developers assume that transcription tools misspell words or make other errors, engineers and researchers said they had never seen another AI-powered transcription tool hallucinate as much as Whisper.

The tool is integrated into some versions of OpenAI’s flagship chatbot ChatGPT, and is a built-in offering in Oracle and Microsoft’s cloud computing platforms, which service thousands of companies worldwide. It is also used to transcribe and translate text into multiple languages.

In the last month alone, one recent version of Whisper was downloaded over 4.2 million times from open-source AI platform HuggingFace. Sanchit Gandhi, a machine-learning engineer there, said Whisper is the most popular open-source speech recognition model and is built into everything from call centers to voice assistants.

So it is not necessarily “Robert from Punjab” not understanding you, it can also be his software. You know those annoying automatic operators asking “What are you calling about?” which interpret “You screwed up my bill!” to “You want to sue a hill?” Yep… but it even INVENTS things.

In an example they uncovered, a speaker said, “He, the boy, was going to, I’m not sure exactly, take the umbrella.”

But the transcription software added: “He took a big piece of a cross, a teeny, small piece … I’m sure he didn’t have a terror knife so he killed a number of people.”

A speaker in another recording described “two other girls and one lady.” Whisper invented extra commentary on race, adding “two other girls and one lady, um, which were Black.”

Imagine a court case where someone’s defense relied on a transcription. Or, the medical transcription software Nabla:

It’s impossible to compare Nabla’s AI-generated transcript to the original recording because Nabla’s tool erases the original audio for “data safety reasons,” Raison said.

Nabla said the tool has been used to transcribe an estimated 7 million medical visits.

AP

Like electric cars… think we may need to dial some refinement time into this.

Category: Science and Technology