It can improve the readability of an article.
It is possible to research a topic using Donald Trump's speeches as long as you are prepared to fact check afterwards.
And yet, they are very good at it. Here's a test I just did:
It's becoming relatively common now, to see job hunting/interview prep advice that advises people to put their CV and the job description into ChatGPT and ask it to write the statement in support of the application. And then ask it to generate a series of interview questions and answers.I’ve been interviewing people at work for a cloud Architect role.
If you’re going to pump the design scenario into chatGPT and present the output at an interview, at least make sure you know what it all means.
“Under DR you’ve mentioned ‘pilot light’, can you tell me more about what that is?”
“Sorry, I left that in by accident.”
Same for “Circuit Breaker Pattern”.
Depending on role we’ve done coding exercises where candidates are given a few days to submit a GitHub repo with an answer and then walk through, live troubleshooting exercises during the interview, presentations, Mural boards etc.
The point is that it is good at giving you factual answers until it stops being good at it. And good luck spotting the difference once you’ve come to rely on it.And yet, they are very good at it. Here's a test I just did:
I asked Google Bard for a list of questions, and then asked ChatGPT for the answers. The first ones were quite easy so I asked it for more difficult ones (question 11 onwards). ChatGPT recalled all of these facts except for the gravitational force one, which is not really a fact, but a calculation. For that, it used one of its plugins and provided the answer that was. It was correct.
In fact, it got them all right except the last one. The organism it mentioned used to be considered the smallest living organism until 2002, but now there is another claim to that title.
I think this is more than accurate enough for most every day tasks, and especially for Welty's blog spam.
This supposes there is an objective “bias” that exists independently from the praxis of the article and its context, and which can therefore be measured and removed to give the “objective, neutral” (i.e., no doubt, centrist) absolute truth. Needless to say, I reject almost every element of that version of the construction of knowledge.I just read about an interesting use-case for ChatGPT, which seems to work quite well: Stripping out political bias in news articles.
I took a Guardian Opinion piece and asked ChatGPT to remove the bias:
I tried to do the same with a Daily Mail piece, but ChatGPT struggled. It tried to explain the article instead. After some more nudges, it got closer, but couldn't do it. Reading the article, it's a very opinionated piece, with few facts, so maybe not the best choice.
Maybe a better use of this would be a tool to highlight bias, rather than try to remove it.
futurism.com
As with virtually all other things that need MOAR POWER, we're just waiting for someone to crack nuclear fusion. The IAEA reckon it could be within the next 25 years.
Well, tbf, 'they' are also working on that.
www.theverge.com
www.creativeboom.com
I'm sure there will be interaction between real people and computer generated material into the foreseeable, but the nature of capitalism is that costs overheads and profits trump everything else, which will result in a serious hollowing out of jobs and human element. Advertising and other short form content is first in the firing line. Some of the wealthy existing ad agencies might dip their toe in as in the link, but new leaner meaner models will come along and upend the marketAI for ideas rather than the end product
Special report: how design agencies are actually using AI in 2024
Forget the hype: how are AI tools actually being employed in the creative world right now? For this special report, we spoke to more than 25 creative agencies to find out. There's been much contro...
I don’t know if this is what you’re after But thought I’d post the whole thread, just in case
