The Artificial Intelligence Problem (or not)

[4859]

Unions run all the LTL shipping in the US. UPS, Fedex, Old Dominion, Estes, etc etc. Everything that would have a 90% immediately implementable AI-driving option, unions are behind the wheel. Unions are also behind the wheel pretty solidly (as in 98% of shipping) in Alaska.

Bro, I love you, and especially what you do but you are literally listing the reasons aspects of your industry are becoming a target.

The unions are so strong in Less than Truckload is because LtL is one of the least touched aspects of the 1980 mca (a loaded deregulation act made in response to a loaded 1930 regulation act) and still have many of its regulations intact, providing among many other benefits, an entry barrier keeping corporations from flooding the market. (On the downside, these entry barriers are making it more difficult to aquire needed new truckers) Only two 'new' big companies I can think of who entered the ltl market since then were ups and fed ex.

The unions used to control a sizable amount of ftl as well. Here's what happened to the trucking industry at large after the excessively deregulating 1980 mca act.

40,000 new shipping companies sprung up between 1980 and 1990, many many many of which of course, were non union. Backed with investor money they engaged in Predatory pricing tactics to undercut their union competition. Union drivers suffered large pay loss as a result. Union membership of the industry at large dropped from 60%, to 20% in just 5 years (1980 TO 1985). A significant portion of which is now left now is indeed ltl.

That's a loss of ground of 40% for the unions. In 5 years.

Ltl has been considered rather niche at 25 billion of the 700 billion shipping market. (As of 2013). However it's a solid, hardworking profitable, sustainable industry that has quickly recovered from the recession and his been making good yoy gains ever since, becoming a key part of the industry, which has inevitably attracted the attention of the shit heads, who see a profitable market to weasel into and harvest.

What is LTL Shipping and How Did it Come About?

(Obviously I'm not posting this to explain what LtL is to a sme like RP, although I'm sure others might enjoy the information, but to support what im saying on the history, and the relationship of the industry to the 1980 mca).


Now I did post all that because I think it's important to look at similar situations that have already happened in recent history.... However, even more pertinent to me than how similar things happened in the past, is what's going on in the present.

If these companies didn't think they could already most likely overcome or sidestep the obstacles you are talking about....

Then why are they investing millions if not billions into making these self driving shipping trucks?

 

[Isiah6:8]

FWIW I feel like worrying about the shipping industry becoming automated is on par with the Yuan becoming a global currency.

If I were going to look at AI and think of what could go quickly, I would look at min wage employees and the jobs they do for corporations and government.

 

[4859]

FWIW I feel like worrying about the shipping industry becoming automated is on par with the Yuan becoming a global currency.

If I were going to look at AI and think of what could go quickly, I would look at min wage employees and the jobs they do for corporations and government.

There are lots of areas, shipping is only one.

Amazon has ready gotten rid of tons of low wage workers in their warehouses with automation.

And high paying ones in their now vestigial and basically doneskies retail sales team with automation.

Besos loves not having to pay meatbags so they can feed their family, or get treatment for being sick, or take time off to spend with their families.

 

[Isiah6:8]

There are lots of areas, shipping is only one.

Amazon has ready gotten rid of tons of low wage workers in their warehouses with automation.

And high paying ones in their now vestigial and basically doneskies retail sales team with automation.

Bezos loves not having to pay meatbags so they can feed their family, or get treatment for being sick, or take time off to spend with their families.

Low wage jobs are valued less for a myriad of reasons in a capitalistic economy. By saying they are low wage workers, you are inherently agreeing with the fact that their contribution to the company is less and in one way or another not as important as others since we are in a capitalistic society. To my earlier point, it shouldn't come as a surprise that low wage work is usually the first to go.

It isn't about making things personal, we in the US are a capitalistic economy. Bezos is also legally required to try and maximize shareholder value. If the employee or employee base added more value than the machine, the machine wouldn't be an option.

 

[4859]

Literally true. Amazon's human retail team got their asses kicked by those algorithms. Entire teams of 15 people outperformed by 2x + by a single employee overwatching a marketplace algorythm.

Low wage jobs are valued less for a myriad of reasons in a capitalistic economy. By saying they are low wage workers, you are inherently agreeing with the fact that their contribution to the company is less and in one way or another not as important as others since we are in a capitalistic society. To my earlier point, it shouldn't come as a surprise that low wage work is usually the first to go.

It isn't about making things personal, we in the US are a capitalistic economy. Bezos is also legally required to try and maximize shareholder value. If the employee or employee base added more value than the machine, the machine wouldn't be an option.

It's more than that. It's definitely not personal, but that doesn't matter in the slightest.

It's basic business strategy. Those low wage workers are not just overhead, but in a well regulated (well regulated doesn't necessarily mean lots and lots of regulations, but well written ones that protect and enable the people to participate in the system) well functioning capitalist system, those 'low paid valueless' workers are potential future competition. IE the American dream. If they get paid enough to not just survive, but build capital, they can start their own businesses and become competition, a threat to the incumbent businesses. So threats are mitigated by manipulating Entry Barriers, and profits are increased by reducing overhead.

 

[policemedic]

Looking at the motor carrier act of 1980, and how it literally slashed union membership in half, I'm guessing another shit ton of deregulations, and another devastating loss for the unions.

Unless we finally manage to start kicking corporate money out of our politics again.

I’m generally not a union guy, so I don’t see union losses as a bad thing.

 

[BloodStripe]

US military's 'Jetson' laser can ID your unique heartbeat hundreds of feet away | Fox News

Facial recognition is so 2008. It's fairly easy to disguise your face with the help of prosthetics and make up. Changing your heartrate on the other hand will require medicine and potential long term health effects. AI will allow for all of our heart rates to become easier to figure out who is who.

This article is one of many reasons why I dislike how transparent our acquisition process has become.

 

[Board and Seize]

Resurrecting this thread to share OpenAI's latest madness.

They've developed a variant of GPT-3 that can generate images from a text description. It is called DALL-E, a deliberate portmanteau of Dali and WALL-E.

Here's the white paper.

If the board is interested in this general topic / subset of "AI", I'd be happy to nerd out on GANs, etc. Ask me about AI Dungeon or Talk to Transformer.

 

[Florida173]

I've done a little GANS with deepfakes. I've been wanting to look at screwing with autonomous vehicles. My wife was working on some of that with work. There was some interesting topics at the car hacking village 2 years ago

 

[Board and Seize]

I've done a little GANS with deepfakes. I've been wanting to look at screwing with autonomous vehicles. My wife was working on some of that with work. There was some interesting topics at the car hacking village 2 years ago

I was considering dropping a white paper at Hack the Sea Village for 27... I'm pretty jelly that you seem to be a regular attendee!

As for vehicular autonomy:

Udacity has a nice looking nanodegree track for self-driving cars.

And then, if you just want to hop in and get your hands dirty, there's all the open source diy projects/communities out there, many influenced or even outright started by Chris Anderson, of 3DR and Wired fame.

• In the air,
  • (software)
  • (hardware)
  • (3DP bodies)
• on land,
  • (or with robots)
• and sea.

 

[Florida173]

I was considering dropping a white paper at Hack the Sea Village for 27... I'm pretty jelly that you seem to be a regular attendee!

As for vehicular autonomy:

Udacity has a nice looking nanodegree track for self-driving cars.

And then, if you just want to hop in and get your hands dirty, there's all the open source diy projects/communities out there, many influenced or even outright started by Chris Anderson, of 3DR and Wired fame.

• In the air,
  • (software)
  • (hardware)
  • (3DP bodies)
• on land,
  • (or with robots)
• and sea.

My wife and I go anytime we are stateside. I was deployed for 26 and obviously safe mode this past year.

Did you see the talk on spoofing AIS?

 

[Board and Seize]

My wife and I go anytime we are stateside. I was deployed for 26 and obviously safe mode this past year.

Did you see the talk on spoofing AIS?

I have yet to make it in person, but I watched as many yt's as I could find and had time/interest for. This one is pretty scary. In fact, the fragility & vulnerability of our entire maritime transportation system is damn scary.

Thank goodness we've got a fair few whites and greys out there thinking about this stuff too.

 

[frostyred]

If the board is interested in this general topic / subset of "AI", I'd be happy to nerd out on GANs, etc. Ask me about AI Dungeon or Talk to Transformer.

Oh man, I saw a video for AI Dungeon earlier and was laughing very loudly.

 

[Kaldak]

Do we want a specific thread for this topic @Board and Seize ? I'm very lost, so I'll leave it you.

 

[Florida173]

Oh man, I saw a video for AI Dungeon earlier and was laughing very loudly.

All my adventures ended up in crazy alien sex thrillers

 

[Board and Seize]

Do we want a specific thread for this topic @Board and Seize ? I'm very lost, so I'll leave it you.

Haha, well, the previous exchange between @Florida173 and myself was a bit of a sidetrack to the intended topic. I'm happy to keep AI-ish stuff going here, and also to decipher some of the jargony stuff.

Back to the top of me resurrecting this thread. You're likely to have had news of various "AI"*
text generators popping up in your feed over the past two years or so.

The first big one that hit the news was GPT-2, which when given a text prompt would generate new text to follow. This is what you can interact with at Talk to Transformer. This uses a specialized kind of ML (Machine Learning)** data structure called a Neural Net (NN). One particular kind of NN is the Generative Adversarial Network (GAN). GANs are found across almost the full spectrum of current and cutting edge AI, from computer-vision object-detection and -defeat (all those captchas you do are your contribution to creating/munging the datasets used to train these networks) to twitter bots, deepfakes and more.

Though perhaps unintuitive, prediction and detection/recognition are fundamentally the same task - just with opposite valence (as an analogy, speakers can be used as microphones - in this case sound production and detection). So you take two NNs. Train them on the same dataset (you show them a bunch of examples of 'right' and 'wrong' at whatever task - say recognizing objects in photos, as with the recent captchas). Then you set them against each other. You task one to predict/produce, and the other to filter/detect. Then you recursively train them on the results of that. As a result, they get successively better. This can be imagined as any of the OpenAI Alpha___ game AIs. That's more approachable, because you just set them to play against each other and record the results, then fast forward and have them do this thousands, millions, or billions of times. The results of the 'learning' is basically an insane number of variables (in the algebraic/cs sense) and weights for each of them.

Okay, back to GPT-n. So this is a NN trained on a huge corpus of basically all reddit comments from the top many subreddits over a period of years. This is a devastating amount of data. Then the model has some large number of 'parameters'. This is basically the variables and weights - and these are pretty blackboxed in that a human can't really penetrate what they 'mean' or correlate to.

GPT-2 dropped in early 2019 with ~1.5 billion parameters. This was the big splash in the news, and they didn't actually release the full model for fear of what shenanigans would occur. They released successively 'larger' models (more params) and then we started seeing stuff like TtT and AI Dungeon.

Then in May of 2020, GPT-3 came out. Basically a much larger, more refined version. It's up to something like 175 billion params now. Since the model is effectively just these params and their relationships (it's a database) it's way too large to run on typical commercial home equipment. Most of your ways to interact with these models (AI Dungeon and TtT again, as well as others) utilize cloud computing resources and provide you with a terminal via webpage.

Phew!

So the new thing that those OpenAI maniacs just dropped today is an evolution of these text generation/'understanding' AIs to create new images - as in draw them. You can write a ranndom description like "sketch of a russian dancing bear with a scorpion tattoo" and it will generate an image that matches. And even from this first glimpse, it is shockingly good.

This took me entirely too long to respond, sorry.

*I get really pedantic over the term Artificial Intelligence, and hold a strict definition. This is othertimes called Strong AI or General AI or True AI. These terms are in reaction to marketing bs that calls every instance of ML AI. To be clear, there is no known extant instance of AI. It is a future possibility.

**Machine Learning is a more accurate umbrella term for all of the crap that gets called AI today. That just means any program that uses any one of a wide range of statistical analyses (typically regressions) to separate signal from noise or pull useful info out of a heap of uninteresting or confusing info.

edit: Oh yeah AI Dungeon! This uses the GPT-2 (or GPT-3 if you pay for it) model to act as a reactive text-based story-teller or game master. It gives you a prompt, and then you start writing, narrating your actions/reactions/thoughts/etc. And then AI Dungeon tells you what happens next. It is freaking awesome, but it isn't perfect, and garbage-in-garbage-out. It can get stuck in a loop, and if you nudge it towards teenage fantasy, it will run with it (all that reddit training yo!).

 

[frostyred]

It can get stuck in a loop, and if you nudge it towards teenage fantasy, it will run with it (all that reddit training yo!).

Hey, at least it wasn't 4Chan training...

 

[Board and Seize]

Well boys and girls, worry not. We can melt the AI brains by dancing!

(slow build/burn on this one, if you're into this stuff, it's totally worth a couple minutes of your time)

 

[Florida173]

Well boys and girls, worry not. We can melt the AI brains by dancing!

(slow build/burn on this one, if you're into this stuff, it's totally worth a couple minutes of your time)

Have you played much with openCV? I have my kit coming from a kickstarter campaign soon.

OpenCV AI Kit

 

[Board and Seize]

Have you played much with openCV? I have my kit coming from a kickstarter campaign soon.

OpenCV AI Kit

Not really, I've poked at it a bit, and I messed a bit with some open Kinect packages for 3D scanning. I need to stop watching other folks get their hack on, and do some myself!

Any plans for your kit? I've only just cracked the lid on the hardware scene, with Arduino and Raspberry, but it's amazing how deep you can go these days with 'custom' off the shelf boards and components. I'd love to see some results from your experimentation!