It is even controlling my clutter. I was watching Guys and Dolls on Prime while cleaning out bankers boxes of files, when in an unrelated folder I came across the cast photo from when my son was in a community theater production of it.

Not creepy at all!

The internet is the only place where yesterday never disappears. With AI in the mix and our smart gadgets secretly listening to us, it feels even creepier. That’s why I think twice, sometimes three times before saying, posting anything publicly or clicking on ads or links.

Instead of boosting search engine optimization, people can now create their own content to boost their products and it is spit out as some kind of packaged up "truth".

Remember back when you were in Jr High and would pass notes to your best friend? My mom always told us "Don't put anything in writing that you wouldn't want on the front page of the newspaper." She was right. We just now share everything over the internet and allow the internet to listen to everything we say, record every purchase, know intimate details of our lives.

If you don't want the world to know what you are saying, don't say it.

I am less concerned with AI knowing what I say than I am with AI replacing people in the workforce. We all get upset with the clumsy customer service bots who can't understand or answer anything correctly, but they are juvenile AI applications. I see how we are using AI in data analytics and data science. In 10 years my industry will be done by AI. Beyond the initial discovery call where I find out exactly what the group that needs info actually needs AI can write the requirements, write the code to extract the data and create a report. It doesn't do everything right, it is still learning, but it will learn and my industry will be as useless as elevator operators.

Here are the top 40 jobs that are expected to be eliminated by AI:

• Customer Service Representatives
• Sales Representatives of Services
• Market Research Analysts
• Management Analysts
• Data Scientists
• Public Relations Specialists
• Technical Writers
• Editors
• Writers and Authors
• News Analysts / Reporters / Journalists
• Interpreters and Translators
• Proofreaders and Copy Markers
• Web Developers
• Public Safety Telecommunicators
• Business Teachers, Postsecondary
• Economics Teachers, Postsecondary
• Library Science Teachers, Postsecondary
• Political Scientists
• Historians
• Mathematicians
• Statistical Assistants
• Demonstrators & Product Promoters
• Models
• Hosts and Hostesses
• Concierges
• Advertising Sales Agents
• New Accounts Clerks
• Counter and Rental Clerks
• Telephone Operators
• Ticket Agents and Travel Clerks
• Broadcast Announcers & Radio DJs
• Brokerage Clerks
• Farm & Home Management Educators
• Telemarketers
• Personal Financial Advisors
• Interpreters / Translators (duplicate reference in original)

Don't send your kids to college - send them to a trade school. AI isn't able to do construction, change a tire, or give someone a massage.

Jennifer, yes. We need plumbers and electricians!

" Here are the top 40 jobs that are expected to be eliminated by AI: "

Pipe dreams and wishcasting. At least in the next little while.

It will be a long time before many of those jobs will actually be replaced by AI or robots.......For some reason, techbros want us to think they've got an awesome product in generative AI, but the truth is that fewer and fewer people are convinced that it's worth what it costs.

The accuracy and precision is nowhere near ready to be fully employed and replacing most human beings at work, so the idea that it could teach or sell something (without human intervention) is absurd. Maybe in 25 or 30 years. Maybe. Its ability to create visuals of real people is getting better, faster, so I can see it could replace models, announcers, reporters a lot sooner. But not anything that requires new ideas or information - remember all it does is synthesize information that's already available publicly on the internet, and it lies a lot. People are beginning to realize that.

I was curious how AI collects data and thought why not ask the source.

What it says:

*******************************

1. Large training crawls (infrequent but massive)

When companies train a new AI model, they typically collect huge datasets by crawling the web at scale, including forums, blogs, and social media.

• Datasets like Common Crawl periodically scan large portions of the public web and are used in many AI models.
• Common Crawl itself runs roughly monthly crawls of billions of pages.
• These datasets then become part of training corpora containing trillions of words from forums, reviews, and websites.

Typical cadence:

• Large web crawls: about monthly or periodic
• Model training using that data: every few months to years, depending on the model.

So an AI model might indirectly include forum posts that were crawled months or years before the model was trained.

2. Retrieval crawlers (continuous / near-real-time)

Some AI systems also run ongoing crawlers that fetch content from websites to answer user queries.

• These bots continuously crawl websites and forums to retrieve up-to-date information.
• Traffic from such AI retrieval bots has increased rapidly in recent years.
• Industry data suggests hundreds of millions of bot scrapes per quarter across websites.

Typical cadence:

• Individual pages might be revisited every few hours to weeks, depending on popularity and update frequency.

3. How often a specific forum is scanned

It varies widely:

Forum size/type

Typical crawling frequency

Major forums (Reddit, Stack Overflow)

Minutes to hours

Active niche forums

Daily to weekly

Small/rarely updated sites

Weeks or months

Large platforms sometimes sell or license their data directly to AI companies instead of relying purely on scraping.

*******************************

But AI is doing more than just synthesizing info on the internet. AI, along with robots have been used in hospitals and pharmacies to improve productivity, and jobs are not replaced by them.

Trade schools are great - HVAC technicians I know make a very good living. But many other professionals, such as doctors, pharmacists, surgeons, and engineers, cannot be replaced by AI.

I don't trust AI now and knowing how they become is even more reason not to trust them. How do they determine what is fact and what is opinion?

Have you watched Hidden Figures - what happened to the human computers?

I joined the workforce just as PCs were introduced to small business. Many jobs have disappeared since then. Remember when we had a secretary take notes with shorthand and transcribe them by typing the notes? How about the person who kept the ledgers and manually figured out how money was being spent? You needed a person to answer the phone and manually set appointments. . .

When I started working in Real Estate the MLS was still a paper directory delivered to every RE office. Each office had to turn in their listings when the Title Reps brought the new MLS to your office, which contained all the information from the previous week.

What happened to all of those jobs now that everything is entered into a database by the realtors? The typesetters, the printers, the Title Reps who delivered the books and got the updates. There were two stories in a large office building dedicated to producing and delivering the MLS for San Diego County. Bright MLS which covers most of the Mid Atlantic Region has less than 200 employees. California MLS employs less then 100 people.

When was the last time you had film developed?

Do you still use a travel agent to book a flight or hotel?

How often do you go into the bank and have a teller assist you instead of just using your ATM Card?

How long did it take for video rental stores to disappear once streaming was available?

When was the last time you relied on a car salesman to tell you about the cars that are available? When I purchased my last car I did all the research online, found out that a car was being produced by the manufacturer that had all of my preferences, so I contacted the local dealer and asked them to have it delivered to their facility so I could pick it up.

Think about car mechanics - used to be that they had to understand how the mechanics all worked and had to be able to diagnose what was wrong with your car. Now they attach a computer to the car and are told what needs to be fixed and how to fix it.

People who are in their 40s or 50s maybe safe, but our kids - those in their teens, twenties or thirties are going to see these fields disappear.

I am already seeing it within my area of technology.

I use AI to promote and track requests, create transcripts and meeting notes, create requirements from those transcripts and assign follow up and work tasks. 2 years ago I did this manually.

If your child has the opportunity to learn subjects based on their current knowledge and taught based using the method that best suits their learning style and paced based on their ability to master new knowledge rather than be part of a 30 student class where the teacher has to teach based on the how quickly the entire class learns the information and teaches the way the majority learn, why would you not want individualized teaching based on what works best for your child.

If your job can be automated it will be automated and the technology is advancing very quickly.

@palisades_ " But many other professionals, such as doctors, pharmacists, surgeons, and engineers, cannot be replaced by AI. "

Radiologists are going to be largely replaced by AI, but most doctors will be safe for the immediate future.

Pharmacists very well may be replaced, automation will take over dispensing tasks and AI is already being used to evaluate potential drug interactions and used to identify issues with compliance (are people getting their refills on time, indicating they are taking their meds regularly). My company monitors Medication Reviews and Medication Adherence for 10 million Medicare recipients. We send the information to the pharmacy when a prescriptions is entered into the system to be filled. Pharmacists deliver the information, but they are no longer required to know the information - AI feeds it to them.

Attorneys in the courtroom won't be replaced, but AI is already automating document production, document review, and legal research that is currently done by paralegals and junior associates.

Again - if your job can be automated - it will be automated and far more can be automated than you would think.

A google search was never the end all be all for any type of research. Google AI is a fun toy for the masses to play with. AI being used for business application is targeted and much more accurate. Sure it still has room to grow, but it has grown fast and does eliminate a lot of work tasks. That means people without jobs.

Anyone who thinks AI will take over any time soon hasn't been to the grocery store recently. How many times does the machine get it wrong? Wrong prices, triggers that things are in your cart, when all it really sees is the cart... AI is stupid. It has a very long way to go. It's a tool to be used by many folks, but only as good as the program that runs it.

Become a doctor, they'll always be in demand, and ain't no one going to a robot for healthcare.

Robots are only as good as the programmers. And it takes knowledge of that profession for it be programmed correctly. So fewer jobs, maybe. Gone? Unlikely.

Pharmacists very well may be replaced, automation will take over dispensing tasks and AI is already being used to evaluate potential drug interactions and used to identify issues with compliance (are people getting their refills on time, indicating they are taking their meds regularly).

…….

Pharmacists deliver the information, but they are no longer required to know the information - AI feeds it to them.

The problem with AI is that it can only generate information based on the data it was trained on. Pharmacists are the final checkpoint before any medication is dispensed to patients. While AI can help provide drug information and identify potential drug interactions, only through direct interaction with patients can pharmacists learn what else they may be taking or any new conditions they may have developed that are not yet documented.

In addition, monitoring disease progression requires strong communication with other healthcare providers, such as doctors and nurses—something that is currently beyond AI’s capabilities. Finally, many patients prefer speaking directly with their trusted pharmacists when discussing medications and concerns. For these reasons, pharmacists are not replaceable by AI.

Become a doctor, they'll always be in demand, and ain't no one going to a robot for healthcare.

Exactly. Same for pharmacists and surgeons, although AI tools can assist them to do better.

What was once dismissed as useless often becomes essential, proving that progress rarely arrives without resistance.

Telephones

Electric light bulb

Cars

Airplanes

Computers

Televisions

The Internet

Cell Phones

AI

I don't think we're saying AI won't take the place of many things. I think we're saying that it is decades away from being able to replace very critical jobs. Driverless cars are hugely promising. Right now, they're worse at driving than humans in some critical ways. I expect that to change, but it's not there yet.

AI is a very long way from being usable. Great tool, in the right hands.

" How do they determine what is fact and what is opinion?"

Exactly the problem. Or a worse problem - how to they determine what's correct and what's a lie? The frequency of appearances on the web is not indicative of truth.

There's also a big problem of stolen copyrighted data/information, and the appropriation of people without their consent. Grammarly is the most recent culprit:

'Writing tool Grammarly has disabled an AI feature which mimicked personas of prominent writers, including Stephen King and scientist Carl Sagan, following a backlash from people impersonated.

The Expert Review function, which offered writing feedback "inspired by" the styles of famous authors and academics, was taken down this week by Superhuman, the tech firm which runs Grammarly.

The feature was met with resistance, including a multi-million dollar lawsuit, from writers who found their names and reputations used as "AI personas" without their consent."

What I am saying is that your experience with google AI and chatGPT is not a good example of industry use. It is being applied more often than you realize, is working well and we are already seeing job displacement that has been attributed to AI.

You are the same as the people who insisted the electric lightbulb would never replace the tried and true oil lamps.

I will give you that it is probably useful in many areas but how do we know when it has made a mistake?

Interesting thread. Yesterday was the anniversary of the ”Luddites”. March 11, 1811

https://www.ebsco.com/research-starters/history/luddites-destroy-industrial-machines

https://www.smithsonianmag.com/history/what-the-luddites-really-fought-against-264412/

“On March 11, 1811, a band of armed English workers attacked a stocking factory in Arnold, near Nottingham. Claiming to be under the direction of a leader named “General Ludd, ” they proceeded to destroy knitting machines, or frames, of a new type designed to produce wide fabric for cheap seamed stockings. Through the months that ensued, their action was repeated in forays of escalating scope and violence, first in the hosiery shops around Nottingham, then in the woolen mills of the West Riding of Yorkshire, and finally in the cotton mills around Manchester. Slow to react at first, the British government eventually dispatched some thirteen thousand troops to restore order. By December of 1812, the main wave of frame-breaking had subsided, partly because of vigorous suppression and partly because of improved economic conditions. However, isolated incidents of industrial sabotage by Luddites continued to occur until 1816, and Luddism—in the more general sense of violent opposition to technological change—experienced a resurgence among British agricultural workers in 1830.”

" What I am saying is that your experience with google AI and chatGPT is not a good example of industry use. "

Actually, it's not just my experience, it's my reading of what other people - in various industries and professions - are saying.

" You are the same as the people who insisted the electric lightbulb would never replace the tried and true oil lamps. "

Nope, because I never said "never". I'm sure (and I hope) that it will get good enough to do the jobs it's being touted for, but right now it's not there. And, given the speed of changes we've seen, I think it's not going to be there for 3 or 4 decades. I'm not against generative AI in general, I'm against using it unsupervised for tasks and in situations that it is neither accurate or precise enough to manage properly. I'm not against electricity, I'm against uncontrolled electricity.

@Elmer J Fudd - this is twice in one week that you and I are in agreement. What is happening to my reality?

Indeed, some very positive signs. Likely there's much more that will come along.

My bookclub recently read a book on AI. We didn't want something overly technical but wanted to know the impact of AI. Here is a summary of it. AI is coming faster than we realize.

"Drawing on more than 40 years of work on AI, this book is for the general reader as well as for professionals. It demystifies artificial intelligence, focusing on its impact rather than on how AI systems work. It addresses the most challenging issues that are currently being debated – the business, risk, legal, ethical, management, and philosophical implications of this remarkable technology.

AI is 75 years old. In the early decades, breakthroughs in AI – in the technology and its application – came every 5-10 years. They are now coming every 6-12 months. In this era of unprecedented technological advance, it is likely that the most impactful systems have yet to be invented. The book claims that balancing the benefits and threats of artificial intelligence – saving humanity with and from AI - is the defining challenge of our age."

I think I've shared a story of an early expert system from long ago - maybe the 1980s. Forget trying to verify details, my memory is general only and they don't matter. An abridged version:

An engineer named Fred managed water releases from PG&E’s Sierra reservoirs (I think) , balancing many changeable factors including hydroelectric power generation requirements, agricultural water supply needs, recreation (white water rafting), snowpack and rainfall levels past and expected, and future demands. The job was complex, and Fred’s expertise was essential throughout the runoff season. He could only vacation in winter.

As his retirement approached , no replacement had been trained. Instead, an expert system was developed during his final two runoff seasons. Students followed Fred daily, asking him to explain his decisions, alternatives, and expected consequences.

The expert system, which was named "Fred", captured his expertise and was tested during the first season after completion, with Fred delaying retirement to oversee it. It worked successfully and became the tool used to manage the complex task of the reservoir release system.

As before, I think this happened during the 1980s. It was certainly in the 20th century but again, my recollection is general.

I cannot stop laughing 🤣 The person who would lecture us all on how we need to be seen by the most expert of medical personnel, is trying to say a robot would be okay. Unbelievable!

I don't think there's a person on this thread that didn't realize that AI was being used. But I think it's a far jump to say it can do any and all jobs better than people. Doesn't matter how long it's been around, it matters how well it's being programmed. And that is still not happening. And it will still take professionals to do the programming. So those jobs are not "going away". No matter how much bluster and how many long posts, no one said that AI hadn't been being used. I'm pretty sure the thought has been floated that it could be a great tool. There's a lot of gray area in between always and never.

🙄

" I don't think there's a person on this thread that didn't realize that AI was being used. "

Only you, it seems.

Failed effort at deflection. You completely misunderstood my comments.

Delusional much?

From my own post

"Anyone who thinks AI will take over any time soon hasn't been to the grocery store recently. How many times does the machine get it wrong? Wrong prices, triggers that things are in your cart, when all it really sees is the cart... AI is stupid. It has a very long way to go. It's a tool to be used by many folks, but only as good as the program that runs it."

Unless you didn't figure out that it was already using AI at the grocery store. I mean I sure gave an example of current usage. Are you serious? Ha! 🤣 You aere just making me laugh with your completely obscure ideas. You don't think any one of us has seen what you've posted before, or nobody can see what I posted in this thread? That is so odd.

You're all too predictable. You've had a number of comments in the past few days that I and others have disagreed with (factually) as in this thread. You're trying to say "no, that wasn't what I meant" as you've done in the past and retaliate with personal insults. No need but be sure others know and recognize your familiar pattern.

You can continue on your own with this if you want to. I won't reply.

rob, your still thinking old school. I have been a programmer, report developer, data architect . . . I fully understand how we use big data under human supervised programming.

With AI and ML (Machine Learning) and DL (Deep Learning) where a person isn't programming the computer to do some kind of specific analysis - the computer is taking all the data available and programming itself as it discovers patterns, continually improving as it gains more and more information with virtually no human intervention. They will find information and patterns and connections that we never even thought to look at.

OMG, we need a hilarious emoji to respond to this, Jennifer!!!!!

I often say please and thank you to Siri. She usually responds with "you're welcome" and "my pleasure". I will question her further. I have asked silly questions like what is your favorite color or when is your birthday. Sometimes she responds with a real answer and other times she says she cannot answer that question. She can be entertaining. But I think that my expose my having too much free time!!

With AI and ML (Machine Learning) and DL (Deep Learning) where a person isn't programming the computer to do some kind of specific analysis - the computer is taking all the data available and programming itself as it discovers patterns, continually improving as it gains more and more information with virtually no human intervention.

That description is somewhat inaccurate because the system does not truly program itself; the learning algorithm is still written by humans. In machine learning and deep learning, humans design algorithms that allow computers to learn statistical patterns from data rather than following explicit rules. During training, the model adjusts its internal parameters to improve performance and may discover useful patterns that humans did not explicitly program. This happens because humans tend to think in concepts and categories, while AI networks operate in high-dimensional mathematical spaces, allowing them to detect patterns across a much broader range of variables. AI programs itself is a myth.

I can speak for Boeing in that they are using AI extensively. And if Boeing is, Airbus is too. Boeing had already been using AI when I retired in 2020. I worked in engineering and saw a lot of lower tech jobs disappearing.

Here's an article that is kind of techy but will give you an idea of what Boeing is doing.

Boeing’s AI looking in the right places? https://share.google/tUinN5T47Phe3xNWG

I don't think AI is "decades" away either. It's happening now quite fast. Read up on how the job market will look 5 years now.

The March issue of The Atlantic had this as their cover story. My take away from reading it is it’s sooner than you think.

I’m not AI. But I tried Hush Glow and wasn’t impressed. The sponge application isn’t a good idea IMO.

What did AI say about the applicator?

AI isn't decades away. PROPERLY working AI is a long way off. And not all jobs will go away because of it.

^^^ tell that to Boeing and its thousands of suppliers. Along with other tech companies.

Microsoft claimed a YEAR ago that 30% of its code is written by AI. Probably more today.

https://www.cnbc.com/2025/04/29/satya-nadella-says-as-much-as-30percent-of-microsoft-code-is-written-by-ai.html

This is a headline today.................

Republicans release AI deepfake of James Talarico as phony videos proliferate in midterm races

I don't care which party does it. It's wrong. Do you really think AI is still a good idea?

AI?....I ya ya AI !. So, we decide to create a "computer" that is way ahead of us in solving many of today's problems. One (or several) AIs being created across the planet which can outthink us many times over (as intended) and perhaps even answer questions we've haven't even thought of yet, it's not too far off, it's right around the corner. Our main challenge will be to make certain that AI doesn't outsmart us even though the main idea is to make it smarter than us. I'm sure this will end well, nyuk.

^^ Oh, I am utterly convinced this is not going to end well...

The problem here is it is new and people are afraid of change. It happens with every technological advancement.

You have choices - you can embrace the change and learn as much as you can about AI and understand the pros and cons and how to use it well or you can resist change and push back and argue against it and declare doomsday or ignore change and pretend it isn't happening.

Regardless of your response it is not going away, it will become the new norm and you will either be prepared or not.

Think of how automation changed the factories and how the steel mills closed as other countries implemented automation ahead of us. It was painful, lots of people lost jobs, but we muddled through and came through to the other side. We still hear lots of people bemoaning the loss of factory jobs and wanting to bring factory jobs back. We are bringing factories back, but the number of jobs is a fraction of the old days because machines have taken over.

Think back to when their was a secretarial pool that supported management and executives. Computers became the norm. I still know some older people who won't use e-mail or text messaging - they may still have a landline phone. They have resisted change, but now find there are things you just can't do anymore with just a phone. You are required to use a computer to set up an appointment or to reach out for customer support.

Pick your path, but understand where that chosen path may lead and give up on the idea that this is something that may not happen. Change and progress don't stop because people resist.

My main problem with it is that it can be used as a tool for evil and it is almost impossible to see that coming. We already have to deal with lying and dishonesty on a major scale and AI isn't going to be that obvious.

And being required to use a computer, well that leaves a lot of people out in the cold. Either the elderly or the less fortunate are out of luck. Even now, most shopping is done online if you really want a choice. Brick and mortar stores are going away and the ones remaining carry little stock and choice.

" The problem here is it is new and people are afraid of change. It happens with every technological advancement. ""

No, the problem here is that it's new and stupid, and most people are afraid of change. But even many of those who embrace change know that it's still too problematic to be used for anything important (and certainly not unmonitored) - ask the people arrested or sent to prison erroneously because of AI. I wonder if anyone's keeping track of the medical problems caused for people because of AI errors. One day, it will be smart, but right now - I wouldn't trust it for anything.

Unfortunately, tech companies are pushing it with abandon, and too many people like sparkly, new things so they can look current and avoid looking like old fogeys - and that leads to it being used too fast, too frequently, and in too many inappropriate ways that it's just not ready for.

" it can be used as a tool for evil "

It's a technological tool, like smartphones or GPS trackers......Like all tools, they can be used for good and evil, and it's up to the individual to use it for the right purpose. Even a hammer is like that - a tool that can be used to build a house, or smash someone's head in.

I am not averse to change. I know AI is here to stay and that the technology continues to advance at a rapid pace. I think that anyone who is computer literate or reads/listens to the news who is not aware that AI is becoming more and more embedded into all aspects of pretty much everything has their head buried in the sand.

And I am convinced THIS technological advancement is not going to end well. I don't mean today or tomorrow, next week or even next year....I mean the long-term implications on society, and even humanity, as a whole.

AI isn't perfect - no one is saying that it is, but you have to weigh the good against the bad.

When you say 'Ask the people arrested or sent to prison erroneously because of AI' don't you need to weigh that number against the total number of rightful arrests made due to AI.

This is a study of Live Facial Recognition being used in London and their results after monitoring the process for 1 year.

https://www.met.police.uk/SysSiteAssets/media/downloads/force-content/met/advice/lfr/other-lfr-documents/live-facial-recognition-annual-report-2025.pdf

Some of the highlights:

a. 2077 alerts, of which 2067 were true alerts. 1

b. 10 false alerts from the ~ 3,147,436 faces that passed the LFR cameras. This results in an annual false alert rate of just 0.0003%, of estimated faces seen. Of the 10 false alerts, only 6 were spoken to. None of these engagements resulted in an arrest.

c. 962 people were arrested

d. 25% of all LFR arrests during the reporting period relate to offences involving violence against women and girls (VAWG)