And the only one that wouldn't be pretty easy to replace.
[NA] Practical tariff impact
- Thread starter ArrowFlyer86
- Start date
Hopefully not the same instructor for consecutively scheduled classes…
Half Fast
Touchdown! Greaser!
- Joined
- May 7, 2016
- Messages
- 21,187
- Location
- Central Florida
- Display Name
- Half Fast
- Home Airfield
- KGIF
Hair stylist?
Cap'n Jack
Final Approach
While it is anecdotal, added to other information, it is a part of a whole. Truthfully, I thought it was a shorter-term thing. She is smart not to criticize the Chinese government where others can hear. I think I have pictures of someone's vegetable garden being bulldozed in 2019, but I don't dare post them on-line. I've noted the Chinese government in some "private" wechat conversations.
Here's the link, and it gives a good overview:
One-child policy - Wikipedia
- Joined
- Jun 11, 2015
- Messages
- 5,612
- Display Name
- Piperonca
- Home Airfield
- M52
Amateurs should have known better than bringing knives to a gunfight. Y'all didn't think @Half Fast was carrying heat, did you?
Last edited:
flyingcheesehead
Taxi to Parking
- Joined
- Feb 23, 2005
- Messages
- 28,184
- Location
- UQACY, WI
- Display Name
- iMooniac
- Home Airfield
- KUES/KMKE
The demand for unskilled labor is going to go off a cliff at some point. At that point, we either will do UBI or some other "utopian" solution, or we'll probably have a civil war on our hands.
It's gonna be some kind of topian, whether it's "U" or "Dys" remains to be seen. I'd prefer to at least be thinking about how we can push things toward "U".
I've observed some of those sea changes, and thought about a lot of the older ones a great deal. But, this one feels very different. AI seems to be changing things far faster than ever before.
Because a computer isn't going to make a decision to invest its own time into doing a new thing. The human is still there for the vision. Even in Star Trek, Data had lots of things he didn't understand about humans and probably wouldn't have made a good business owner.
I'm totally with you on most of that, but as I've stated earlier in this thread, humans aren't necessarily driven to create solely to make money... Much more amazing things could happen if we freed ourselves to pursue creative interests that weren't necessarily driven by the almighty dollar.
I think some of these are going to happen because they are conducive to being done in an automated fashion, and some aren't...
Should be reasonably easy to automate; probably one of the biggest challenges will be developing the right appendages and dexterity to manipulate all the various parts.
As long as the horse can be contained and the machine doesn't make crazy noises, this is something that might be automated. However, I would see it as a tool that could be rented/purchased and used by owners on their horses. Getting the horse to where it can be used and keeping the horse calm will still require a human touch.
Almost any sort of training/education, of animals or humans, is likely to require a human. There will be plenty of tools to help the humans (there already are), but this will be human-led.
Wait until the immigrants all get run out of town, these will be automated before you know it.
Definitely automated.
Haven't you seen Robocop?
This is probably going to be a combination of humans and tech.I think there'll be a lot of tools to help the humans with this, but these folks' bedside manner is important... And I am NOT volunteering to be the first one to get a tooth drilled by a robot, even if it is accurate... It doesn't know how much it'll hurt if it screws up!
I think this sort of thing will be shared in different ways. I think that there will be automated tools for high-volume, modern sorts of work, but there will also be some demand for true craftsmen who can do custom work on rare pieces like you mention.
This is an 80/20. If it's for tourists going out in a popular area for the first time, automation could do it. If it's for really experienced folks in desolate backcountry places that aren't popular, humans will do it, aided by tech like AI trail cameras that give them information.
That one's all you.
Funny... But, this will probably be automated earlier than many things on the list. Sex sells... And if you can sell it without it costing you much money or causing too many ethical issues, you'll be rolling in money. They've got a ways to go yet, but there's definitely work happening on it.
A human has to make the decision to mechanize something to begin with, which is why many businesses aren’t as automated as they could be today.
The world (and society) will evolve, but humans understand the need for some type of gainful utilization. There’s a reason the phrase idle hands are the devil’s workshop has been codified for a couple millennia now.
Half Fast
Touchdown! Greaser!
- Joined
- May 7, 2016
- Messages
- 21,187
- Location
- Central Florida
- Display Name
- Half Fast
- Home Airfield
- KGIF
The demand for unskilled labor is going to go off a cliff at some point. At that point, we either will do UBI or some other "utopian" solution, or we'll probably have a civil war on our hands.
It's gonna be some kind of topian, whether it's "U" or "Dys" remains to be seen. I'd prefer to at least be thinking about how we can push things toward "U".
I've observed some of those sea changes, and thought about a lot of the older ones a great deal. But, this one feels very different. AI seems to be changing things far faster than ever before.
Because a computer isn't going to make a decision to invest its own time into doing a new thing. The human is still there for the vision. Even in Star Trek, Data had lots of things he didn't understand about humans and probably wouldn't have made a good business owner.
I'm totally with you on most of that, but as I've stated earlier in this thread, humans aren't necessarily driven to create solely to make money... Much more amazing things could happen if we freed ourselves to pursue creative interests that weren't necessarily driven by the almighty dollar.
I think some of these are going to happen because they are conducive to being done in an automated fashion, and some aren't...
Should be reasonably easy to automate; probably one of the biggest challenges will be developing the right appendages and dexterity to manipulate all the various parts.
As long as the horse can be contained and the machine doesn't make crazy noises, this is something that might be automated. However, I would see it as a tool that could be rented/purchased and used by owners on their horses. Getting the horse to where it can be used and keeping the horse calm will still require a human touch.
Almost any sort of training/education, of animals or humans, is likely to require a human. There will be plenty of tools to help the humans (there already are), but this will be human-led.
Wait until the immigrants all get run out of town, these will be automated before you know it.
Definitely automated.
Haven't you seen Robocop?
I think there'll be a lot of tools to help the humans with this, but these folks' bedside manner is important... And I am NOT volunteering to be the first one to get a tooth drilled by a robot, even if it is accurate... It doesn't know how much it'll hurt if it screws up!
I think this sort of thing will be shared in different ways. I think that there will be automated tools for high-volume, modern sorts of work, but there will also be some demand for true craftsmen who can do custom work on rare pieces like you mention.
This is an 80/20. If it's for tourists going out in a popular area for the first time, automation could do it. If it's for really experienced folks in desolate backcountry places that aren't popular, humans will do it, aided by tech like AI trail cameras that give them information.
That one's all you.
Funny... But, this will probably be automated earlier than many things on the list. Sex sells... And if you can sell it without it costing you much money or causing too many ethical issues, you'll be rolling in money. They've got a ways to go yet, but there's definitely work happening on it.
Remember - the question was whether it was economically advantageous for a business owner to pay for automation to replace an employee. Replace, not assist. Even if such a 100% machine were possible, it might not make business sense to use one.
Having offered that reminder, I'll consider just one example you discussed above. I think it's possible that you've never seen a farrier at work, so let's look at just what you're expecting a robotic farrier to do.
In the 100% replacement scenario, my wife opens an app on her phone and schedules a time for farrier service. At the appointed time, the farrier robot arrives, likely flown in by drone but maybe it's a self-driving truck. The machine unloads itself and all necessary supplies and tools, then proceeds to catch the horse in the pasture (HAH!), place a halter and lead rope on it (
) and lead it to the paddock. There the machine puts a lunge line on the horse and lunges it for a while at different gaits to observe how the horse moves, whether there's any lameness, whether it transitions easily between gaits and leading leg, etc.Next the machine leads the horse to the barn and places it in cross ties for the shoeing process. The machine lifts each hoof and examines it for damage, disease, and contours. The machine then cleans and trims each hoof, carving each hoof to provide a good base for the horseshoe. If the horse hasn't kicked the machine into the next county by now, the machine will proceed to the actual shoeing.
Based on its observations of the horse's movements, and relying on its mammoth database for judgement (uh, where's that database coming from?), the machine selects an appropriate style of shoe, and possibly pad, for each hoof. It next proceeds to heat a shoe to a red-hot state and adjust its shape to conform to the hoof. Then, while the shoe is still hot, the machine presses it against the hoof, taking great care not to burn the horse. Finally, the machine drives in a set of nails to secure the shoe to the hoof. This process repeats for the remaining hooves.
Once the shoeing is complete, the machine releases the horse and observes its movements for a while to make sure the horse moves normally. Its job complete, the machine then re-enters its self-driving truck or drone and proceeds to the next appointment.
Of course, the machine contacts my wife via the farrier app a few days later to see how the horse performs after being ridden a few times.
Even if such a machine were possible, I can't imagine it would be more cost effective than human labor.
Last edited:
Half Fast
Touchdown! Greaser!
- Joined
- May 7, 2016
- Messages
- 21,187
- Location
- Central Florida
- Display Name
- Half Fast
- Home Airfield
- KGIF
You seem to be arguing that automation will replace things requiring skilled labor, but the need for unskilled labor will decline. Forgive me for being confused.
It will be a phased implementation. Automation is assisting many jobs today. That trend will continue till eventual replacement.
Imagine harder.In the 100% replacement scenario, my wife opens an app on her phone and schedules a time for farrier service. At the appointed time, the farrier robot arrives, likely flown in by drone but maybe it's a self-driving truck. The machine unloads itself and all necessary supplies and tools, then proceeds to catch the horse in the pasture (HAH!), place a halter and lead rope on it (
Next the machine leads the horse to the barn and places it in cross ties for the shoeing process. The machine lifts each hoof and examines it for damage, disease, and contours. The machine then cleans and trims each hoof, carving each hoof to provide a good base for the horseshoe. If the horse hasn't kicked the machine into the next county by now, the machine will proceed to the actual shoeing.
Based on its observations of the horse's movements, and relying on its mammoth database for judgement (uh, where's that database coming from?), the machine selects an appropriate style of shoe, and possibly pad, for each hoof. It next proceeds to heat a shoe to a red-hot state and adjust its shape to conform to the hoof. Then, while the shoe is still hot, the machine presses it against the hoof, taking great care not to burn the horse. Finally, the machine drives in a set of nails to secure the shoe to the hoof. This process repeats for the remaining hooves.
Once the shoeing is complete, the machine releases the horse and observes its movements for a while to make sure the horse moves normally. Its job complete, the machine then re-enters its self-driving truck or drone and proceeds to the next appointment.
Of course, the machine contacts my wife via the farrier app a few days later to see how the horse performs after being ridden a few times.
Even if such a machine were possible, I can't imagine it would be more cost effective than human labor.
The vast majority of what you describe is just learning how horses work. Today's LLM based AI could learn that if someone were sufficiently interested to train it with the data. And, once that's done, not only is it as well trained as an experienced farrier, but that knowledge is infinitely and perfectly copyable. You don't have to apprentice each robot in a multi-year program to bring them up to speed. Further, once you have hundreds of farrier-bots out there working, they all learn from every single farrier-bots real world experience.
The real problem is the hardware. But it seems like a sucker bet to claim that progress in robotics is going to stop. Even today you can buy pretty amazing robots for low tens of thousands of dollars. So, let's say that the farrier-bot is ten times more expensive. That makes them between a hundred and five hundred grand with todays technology. That's an ROI of 5-10 years compared to hiring human farriers. All with technology that we basically have today. Technology is advancing rapidly. I've repeatedly speculated a hundred year timeline. Seems like betting against the robots will be a bad bet.
Here's an example of what can be done with current hardware, not by programmers, but by musicians. These robots were 'trained' just by pushing them through the motions and having them remember. We are very quickly entering a world where we are only limited by our creativity. And don't worry, the AIs are already helping with that too...
Ed Haywood
Final Approach
Remember - the question was whether it was economically advantageous for a business owner to pay for automation to replace an employee. Replace, not assist. Even if such a 100% machine were possible, it might not make business sense to use one.
Having offered that reminder, I'll consider just one example you discussed above. I think it's possible that you've never seen a farrier at work, so let's look at just what you're expecting a robotic farrier to do.
In the 100% replacement scenario, my wife opens an app on her phone and schedules a time for farrier service. At the appointed time, the farrier robot arrives, likely flown in by drone but maybe it's a self-driving truck. The machine unloads itself and all necessary supplies and tools, then proceeds to catch the horse in the pasture (HAH!), place a halter and lead rope on it (
Next the machine leads the horse to the barn and places it in cross ties for the shoeing process. The machine lifts each hoof and examines it for damage, disease, and contours. The machine then cleans and trims each hoof, carving each hoof to provide a good base for the horseshoe. If the horse hasn't kicked the machine into the next county by now, the machine will proceed to the actual shoeing.
Based on its observations of the horse's movements, and relying on its mammoth database for judgement (uh, where's that database coming from?), the machine selects an appropriate style of shoe, and possibly pad, for each hoof. It next proceeds to heat a shoe to a red-hot state and adjust its shape to conform to the hoof. Then, while the shoe is still hot, the machine presses it against the hoof, taking great care not to burn the horse. Finally, the machine drives in a set of nails to secure the shoe to the hoof. This process repeats for the remaining hooves.
Once the shoeing is complete, the machine releases the horse and observes its movements for a while to make sure the horse moves normally. Its job complete, the machine then re-enters its self-driving truck or drone and proceeds to the next appointment.
Of course, the machine contacts my wife via the farrier app a few days later to see how the horse performs after being ridden a few times.
Even if such a machine were possible, I can't imagine it would be more cost effective than human labor.
I'm guessing there is some petting and soothing talk to calm the horse down also. Bet the shoe-o-matic is great at that.
Humans are more complicated than horses. We already have early AI that can sooth humans. Horses will be easier.
- Joined
- Jun 7, 2008
- Messages
- 27,381
- Display Name
- Bob Noel
- Home Airfield
- KBED
I'd love to see that...
Cap'n Jack
Final Approach
In this example, is there any reason to keep using the procedure you describe? Note that I know next to nothing about shoeing a horse. For example, can someone (something) tell from the wear patterns of the shoes on the horse details that help to choose a new shoe? Machine vision can look for damage, disease, and contours. I've worked with 3 companies so far that are working to automate compound synthesis and purification. I've worked on the purification part, as I've invented an algorithm to calculate purification methods. There is someone who decides what compounds to make, and one can buy into an AI that does the retrosynthesis and then tells how to do the synthesis from commercially available compounds: https://www.sigmaaldrich.com/US/en/...ZW0IQdMkBZZwcXbZMBn9ASITbuvnvtdQVyGwtZ8PT6TrQ . You described how it is done now. Once upon a time, people used oars to move ships (I'm going to skip sail. But sailors got replaced, too). Some of the early steam powered vessels were impractical because the inventor tried to emulate the oar. Using a paddle wheel allowed practical replacement of oars on galleys. The screw propellor looks nothing like an oar but is much more efficient at moving a ship than either an oar or paddlewheel.Remember - the question was whether it was economically advantageous for a business owner to pay for automation to replace an employee. Replace, not assist. Even if such a 100% machine were possible, it might not make business sense to use one.
Having offered that reminder, I'll consider just one example you discussed above. I think it's possible that you've never seen a farrier at work, so let's look at just what you're expecting a robotic farrier to do.
In the 100% replacement scenario, my wife opens an app on her phone and schedules a time for farrier service. At the appointed time, the farrier robot arrives, likely flown in by drone but maybe it's a self-driving truck. The machine unloads itself and all necessary supplies and tools, then proceeds to catch the horse in the pasture (HAH!), place a halter and lead rope on it (
Next the machine leads the horse to the barn and places it in cross ties for the shoeing process. The machine lifts each hoof and examines it for damage, disease, and contours. The machine then cleans and trims each hoof, carving each hoof to provide a good base for the horseshoe. If the horse hasn't kicked the machine into the next county by now, the machine will proceed to the actual shoeing.
Based on its observations of the horse's movements, and relying on its mammoth database for judgement (uh, where's that database coming from?), the machine selects an appropriate style of shoe, and possibly pad, for each hoof. It next proceeds to heat a shoe to a red-hot state and adjust its shape to conform to the hoof. Then, while the shoe is still hot, the machine presses it against the hoof, taking great care not to burn the horse. Finally, the machine drives in a set of nails to secure the shoe to the hoof. This process repeats for the remaining hooves.
Once the shoeing is complete, the machine releases the horse and observes its movements for a while to make sure the horse moves normally. Its job complete, the machine then re-enters its self-driving truck or drone and proceeds to the next appointment.
Of course, the machine contacts my wife via the farrier app a few days later to see how the horse performs after being ridden a few times.
Even if such a machine were possible, I can't imagine it would be more cost effective than human labor.
The end result will probably somewhere short of total automation, but much more than there is now. McDonalds apparently didn't reduce employee count with their app and kiosks, but some of the cashiers are doing other things: https://www.cnn.com/2024/09/20/business/self-service-kiosks-mcdonalds-shake-shack/index.html Some places just have an app without the kiosk. The app is convenient because I usually just get whatever I ordered with no waiting.
- Joined
- Jun 11, 2015
- Messages
- 5,612
- Display Name
- Piperonca
- Home Airfield
- M52
People who work around horses can tell you how unbelievably sensitive they are. https://jumpernation.com/two-bodies-one-heart-horses-and-your-heartbeat/
So, some robot comes buzzing right up and the first thing that horse notices is something missing that machines will never replace.
I've seen horses spooked by a small puddle of water. I can only imagine the first encounter with a shoe-o-matic. It would not go well.
Last edited:
- Joined
- Jun 11, 2015
- Messages
- 5,612
- Display Name
- Piperonca
- Home Airfield
- M52
I made a good living inventing algorithms, and seeing them fielded, so we have something in common.
All for it! Whatever increases accuracy and minimizes time. Those are the goals. But some human things won't be as amenable to automation. Take the local tire shop. Somebody's got to get their hands dirty and the job done.
Cap'n Jack
Final Approach
Do you mean replacing tires? That seems amenable to automaton to me. Jack the car up next to 4 robotic arms and swap the tire(s). I also realize there are aspects that I'm overlooking with running such a shop. Automation will gradually change businesses, perhaps as fast as some will expect, and not as completely. Rob't Fulton demonstrated the commercial potential of the steam boat in 1807, yet there were 2,000 windjammers sunk by U-boats in the first world war, over 100 years later! https://www.hrmm.org/history-blog/media-monday-u-boat-sinks-schooner-in-the-first-world-war-1917
- Joined
- Jun 11, 2015
- Messages
- 5,612
- Display Name
- Piperonca
- Home Airfield
- M52
https://www.discounttire.com/
The foundational document I used on the military/OEM side was MIL-H-46855. The gist of it is do a task analysis first, with plenty of end user representation. Look at the environment, etc. Employ mockups, walkthroughs, and prototypes. What happens is you iteratively allocate functions, some to the human, some to automation. Test, verify, and validate. And with most of the weapon systems I'm familiar with, there is always a human in the loop. Maybe distant eyeballs, but always there.
Don't know if Discount Tire is ready for that yet. They will watch the technology, and make their cost/benefit analysis when necessary.
Half Fast
Touchdown! Greaser!
- Joined
- May 7, 2016
- Messages
- 21,187
- Location
- Central Florida
- Display Name
- Half Fast
- Home Airfield
- KGIF
When someone invents steam-powered horses and eliminates their legs and hooves, your analogy will work better.
BUT.... The question wasn't whether or not a machine could be made to replace the human. The question is whether there is a business case for all the non-recurring R&D, plus the cost of building and maintaining such machines (which will regularly get kicked into oblivion), instead of paying a human employee.
Even if I grant the possibility of such a complex machine (highly improbable though it seems), given all the effort and expense needed to create it I don't grant that there will be a business case that justifies it.
And the same is true for other items on my sample list.
BTW, regarding
In a word, no. Horses are living creatures. Their limbs and hooves change, their weight goes up and down, etc. The purpose for which the horse is ridden changes. The surfaces it walks on change. Weather changes what the horse does and how it moves. Horses become injured and lame regularly and require special trimming and shoeing. And on and on and on....
But the robot will also need to dismount the tire from the rim, select the correct new tire and mount it, then inflate it, and then reinstall it. Possible? Probably, at least within some limited subset of rims and tires and sizes. Cost effective? Doubtful, but maybe.
Cap'n Jack
Final Approach
Cap'n Jack
Final Approach
I'm not understanding why this applies to the conversation.
- Joined
- Jun 11, 2015
- Messages
- 5,612
- Display Name
- Piperonca
- Home Airfield
- M52
Has to do with automation; whether or not to automate. Look at the tasks first. Admittedly complicated example but subject material has some commonality. Any further replies to this conversation will be an automated bot.
Cap'n Jack
Final Approach
A lot of automation is far from formal. I've seen synthesis purification go from manual to entirely automated. It wasn't done by looking at the entire task, but rather replacing the manual glass column with a machine using pre-packed columns and pumps. The decision wasn't to automate, but rather improve the process. The chemist still had to do everything except pack the column. Now we add an auto sampler that loads sample on the column, an incremental change. Next, I can calculate the purification method from the analytical run. Each of these increments is a step towards automation over several years. Now, a computer with a database links the analytical and preparative instruments together and robotics move the sample around. The professionals who used to do the purification are replaced by relatively unskilled people who make sure the solvent containers are filled (the systems now tell them that they need refilled), the waste containers are replaced before they overflow- again, the chromatography systems tell them. This didn't happen overnight, it's just that the individual components got "smarter" and a software person tied them together. There was no formal review process, because it happened over years as chromatography equipment reached the end of it's useful life and was replaced with newer, more capable, products, until someone could see how to take the last step in automation.
The example I mentioned earlier, where one can buy into retrosynthesis AI, is another example. There is very little capital cost, but there is fewer need for skilled chemists. That might be the way we get new drug development back in the USA, since we outsourced those skills so long ago, and it will take 6 to 12 years to start getting them back.
Isn't AI redefining software development now? That is a rather disruptive change and may be more appropriate to your comment.
- Joined
- Jun 11, 2015
- Messages
- 5,612
- Display Name
- Piperonca
- Home Airfield
- M52
Going back to the original intent of the thread— having to do with tariffs, I think the concern is how difficult it’s going to be to start up and build robot factories, let alone make a case for expensive specialized ones. I can see where AI might improve your process development. Anymore, the thought of software development gives me a headache. Retired from that. Thoughts of your job gives me a headache too, so I think I’ll bow out for now.
You're forgetting that horses have a pretty keen sense of smell, 300 million olfactory receptors compared to a human's 5 million.
GaryM
En-Route
- Joined
- Apr 16, 2020
- Messages
- 3,087
- Location
- Was KMMU, now KCOE
- Display Name
- Gary M
- Home Airfield
- Was KMMU, now KCOE
It appears the woman is neighing, to add to the authenticity.
Or a steam pop-off valve just let loose
flyingcheesehead
Taxi to Parking
- Joined
- Feb 23, 2005
- Messages
- 28,184
- Location
- UQACY, WI
- Display Name
- iMooniac
- Home Airfield
- KUES/KMKE
And a human will make that decision as soon as it saves them money (or makes them more) compared to hiring another human.
People are expensive and kind of a pain, and automation is becoming ever cheaper and easier.
Also, it doesn't have to be a 100% replacement. If you have a machine that can't do everything a person can do, but makes that person twice as efficient, you can cut one person down to part time, or you can eliminate half of the people who do that job.
Oh, I have, though more often just for clipping and cleaning than shoeing. And that's why the farrier wasn't on the fully automated list! I don't think a machine can handle a horse.
None of your examples were unskilled, so you seem to agree that those folks will go away?
But half of your skilled stuff can be automated too. So at the end of the day, we're talking massive job losses.
Fascinating.
I can tell you that horses get freaked the f out when you fly over them at 500-1000 AGL in a Citabria. Cows just look up at you and keep chewing.
I disagree with absolutes. I work for a F100 financial services company. We’ve automated a lot of stuff, but getting out of customer service employees is off limits, since talking to a human is core to our business model. We have competitors who choose a different operating model.
Customers have a choice to DIY it or have a helping hand. We target those who want a helping hand in a complex space.
flyingcheesehead
Taxi to Parking
- Joined
- Feb 23, 2005
- Messages
- 28,184
- Location
- UQACY, WI
- Display Name
- iMooniac
- Home Airfield
- KUES/KMKE
True, but I would argue that the human touch is something that can't be replaced by automation at this point.
Years ago, Apple moved their customer service to India like everyone else was doing - I don't think they were a month in before they brought it all back. They recognized that stateside customer service was becoming the premium product.
And, I think there will always be luxury brands that hang their hat on personalized, human service.
If we change that to "specialty" service, I'm with you. Computers will be able to do a better job at luxury than humans. They will know more and be able to tailor the experience to customers better than humans. Their ability to recall customer preferences and history will be flawless as they don't suffer the same memory problems as humans.
However, there is a reason every sci-fi story ever written about this upcoming era has some segment of society that refuses to interface with the machines. Often this is for religious or spiritual reasons. You speculate "luxury". I dunno, maybe... I do agree it's hard to imagine Hermes switching to robots in either their retail outlets or factories. But I do think you're right in claiming that there will be people that only want to talk with people for some reason.
flyingcheesehead
Taxi to Parking
- Joined
- Feb 23, 2005
- Messages
- 28,184
- Location
- UQACY, WI
- Display Name
- iMooniac
- Home Airfield
- KUES/KMKE
Heck, I only want to talk with people most of the time, though that's because usually if I'm calling someone else in to help me, I've exhausted all of the normal troubleshooting steps and I need an expert. Sadly, the machines aren't experts yet.
A year ago you would have had an easier time convincing me that the machines aren't experts. Now, it seems like 70% of the time chatgpt is my first stop rather than something I turn to after exhausting "normal troubleshooting steps"
And I'm clearly not alone...
The normal troubleshooting steps are changing rapidly.
tspear
En-Route
- Joined
- Dec 10, 2010
- Messages
- 4,010
- Display Name
- Timothy
For those thinking there was some 4D chess strategy, and the goal was to isolate China. It is working very well: https://www.dw.com/en/china-set-to-...-eu-charm-offensive/a-72357774?mobileApp=true
And some depressing trend lines: https://www.apolloacademy.com/wp-content/uploads/2025/04/042625-ConsumerandFirms_v2.pdf
The Ag sector is getting killed: https://www.cnbc.com/2025/04/28/trade-war-tariffs-full-blown-crisis-us-farm-exporters-say.html
Tim
And some depressing trend lines: https://www.apolloacademy.com/wp-content/uploads/2025/04/042625-ConsumerandFirms_v2.pdf
The Ag sector is getting killed: https://www.cnbc.com/2025/04/28/trade-war-tariffs-full-blown-crisis-us-farm-exporters-say.html
Tim
flyingcheesehead
Taxi to Parking
- Joined
- Feb 23, 2005
- Messages
- 28,184
- Location
- UQACY, WI
- Display Name
- iMooniac
- Home Airfield
- KUES/KMKE
I'm sure things will be changing rapidly, and I'm guessing I'm going to become even more frustrated with it.
Right now, if I'm going to fix something I haven't fixed before, I'll find YouTube videos about it from a handful of different sources. They'll all cover things that the others don't. Unfortunately, it's also fairly common that one of them will tell you to do something that the others say to not do, or do differently, for very good reasons. I figure out what technique I'm going to use by considering the whys.
AI doesn't give a crap about the whys. It only cares about probability. If you have four amateurs that tell you do to it using method A and one pro who tells you to do it using method B and gives a great justification, AI will tell you to do it using method A. Then, it might just use the method B justification for fun!
Garbage in, garbage out... And as of yet, I don't think anyone is trying to filter the garbage going in.
If I'm starting with chatgpt 70% of the time, the other 30% is for something visual that youtube shows. So, I'm with you on that for sure.
That's just not how it works. Current AI is much more sophisticated than simply counting the number of times an opinion is mentioned.
They absolutely are trying to filter that. I can't really give examples because it would reveal where I work, but at a high level weighting of sources both manually and automatically is hugely important and happening constantly. Heck, the manual part is a constant source of controversy and debate.
StraightnLevel
Cleared for Takeoff
- Joined
- Dec 8, 2023
- Messages
- 1,161
- Location
- Houston & SE Wisconsin
- Display Name
- StraightnLevel
- Home Airfield
- T41, KETB
Optimist - the glass is 1/2 full.
Pessimist - the glass is 1/2 empty.
AI - the glass is January 2.
Pessimist - the glass is 1/2 empty.
AI - the glass is January 2.
Because it was trained using Excel date handling?
