No, I canât see the future, some people are just really that dumb. Itâs like putting a bit of peanut butter in a mouse trap, in a rodent infested home, you know youâll get a mouse, the dog, or a baby eventually.
Why it matters : Regulators and technologists were slow to address the dangers of misinformation spread on social media and are still playing catch-up with imperfect and incomplete policy and product solutions.
Yah there are a few times that chatgppt has been completely wrong in its assessment of a collection of resources - notably - history and health - I guess itâs no more dangerous than relying solely on wikipedia for your information. Itâs a tool.
This is a quote that I use quite often talking to people about problems they are having with their computerâŚIt was a quite common quote in silicon valley.
I say this to my boss every time he wants me to use a shortcut or cut a corner. He never listens, and I always have to spend 3X the time later to fix it after something goes wrong.
I am on the mailing list of an pulp story writer who has published by real publishers and operates as an independent, self-publishing on Kindle and publishing serailized stories on Kindle Vella.
He has a following, and promotes his works.
This past month, he received a small fraction of the revenue he receives from Kindle Vella.
Vella has a revenue pool which is divided among authors based on their percentage of their stories which are viewed.
As one would expect, Amazon is unwilling to share its algorithm or explain why the payment was lower than expected.
Like we Forum users do, he and some other authors followed some bread crumbs, and did a little research.
There has been a huge surge in new listings on Vella and Kindle. These authors believe this surge was in writings generated by ChatGpt or other AI based tools.
I have not reviewed their evidence but immediately believe this to be plausible.
They are hoping that Amazon will take action to block the publication of AI generated works.
I emailed him suggesting that Amazon lacks the AI to do so, will probably try to do it with a blunt instrument, after they get negative PR, and probably punish many innocent authors as a result. And I offered him examples of Amabot actions which have hurt 3P sellers.
This should be a train wreck. But this is America, where people paid to watch staged trainwrecks in the early 20th century.
My wife and I do a crossword puzzle together each morning.
We do the WSJ puzzle on Weekdays and Saturday. The New York Times and/or Los Angeles Times puzzle on Sundays.
We have been aware that AI assistance is used by crossword puzzle creators and has been for about 15 years or more.
It has been clear from our 8 puzzles per week that the affects of AI are more and more pronounced.
The same words (with different clues) are appearing in more and more puzzles.
Even when puzzle authors worked by hand and brain, there were some words which were more common than others.
In last weekendâs NYT puzzle there were at least a dozen words in the parts I have done which had appeared in another puzzle we did in the past week. It is clear the writers are contributing less and the computer lacks the vocabulary and ability to use it that a qualified human puzzle creator used to have.
The NYT provides a bio on its puzzle authors for the Sunday puzzle, and the number of computer geeks listed as authors is on the increase, as are the number of long distance collaborators.
Someday the times might even disclose what the software contributors are for their inferior works.
I have been doing their puzzles with editors starting at Margaret Farrar, through Will Weng, and Eugene Maleska (who taught English at my high school). Will Shortz has always had lower standards, so it is not surprising for him to be a Yea on AI.
John Oliver did a piece on AI this past weekend and it was pretty fun.
AI will eventually displace a lot of workers but I think that there will always be work. The problem is that eventually that work will be âprovide blood for the machines.â
Just read an article on the next generation of AI (it seems Microsoft/OpenAIâs Chat GPT, LLM tech is already obsolete) Kosmos-1 MLLM, can get an IQ test question right 22-26% of the time. Which is apparently higher than the 17% of random chance.
AI isnât even close, like the man once said, âAll we have to fear, is fear itself.â
For more information on Microsoftâs Kosmos-1 Machine Learning Model AI.
Be aware and warn your friends to stay alert to ways that social media and other platforms might be manipulating youâor sellers manipulating products/ads.
I understand your concerns about AI overtaking published works. As with any new technology, there will always be abusers and then government will get involved and pass laws against it, such as with drone technology.
It takes a certain amount of pre-planning to use AI more effectively. For example, your crossword puzzles could be written in AI with instructions to AI not to repeat previous crosswords.
There are AI programs that will check for plagiarism, or rewrite the original AI text into more human-style writing, but the influx of new Kindle books may not be utilizing those tools.
I enjoy working with AI text and AI art, and looking to the future, would especially enjoy AI music.
After warning employees not to use ChatGPT, Amazon is now asking its software engineers to start using an in-house AI tool called CodeWhisperer according to Business Insider today.
Amazon it seems while very much all-in on AI, is taking a hard pass on ChatGPT.
Be aware and warn your friends to stay alert to ways that social media and other platforms might be manipulating youâor sellers manipulating products/ads.