General Automation Thread

[Elheru Aran]

Unfortunately, yeah... the only real question is how many people are going to have to die and/or revolt before the various governments give up and declare that either everybody goes on the dole, or everybody who's not working can suck a fart and die.

[FaxModem1]

Speaking of governments becoming affected by automation, Germany may face some interesting problems down the road.

New Statesman

New Statesman
Search
Menu
Angela Merkel at a robotics factory
PHOTO: GETTY
EUROPE 21 SEPTEMBER 2017
Germany's political stability could be threatened by automation
The country's resistance to populism may be tested by changes to its manufacturing industry.
Sign up to the Staggers Morning Call email *
Sign up to the Staggers Morning Call email
SUBMIT
BY CLAUDIA CHWALISZ

Germans head to the polls this Sunday 24 September. With Merkel set to win a fourth term as Chancellor, it has been dubbed a "sleepy" election – particularly compared to the Dutch and French campaigns a few months ago. Populism, while present, has not taken off to the same extent as in Germany’s neighbouring countries.

In a new Legatum Institute report co-authored with Matthew Elliott, we explore in detail why this is the case, evaluating the historical and economic circumstances as well as social, cultural and political attitudes. In short, support for both the populist Left Party (Die Linke/DL) and for the populist right Alternative for Germany (AfD) has so far been concentrated in former East Germany. At the national level, it has therefore been hard for either party to win more than around 10 per cent of the vote.

However, a longer term trend that might disrupt German politics in future election cycles is automation. With manufacturing making up a large proportion of the German economy, a significant amount of jobs are set to shift between occupational groups. According to the OECD, the portion of jobs at high risk of automation in Germany – 12 per cent – is one of the highest among countries measured.

While the elimination of some jobs and occupations does not necessarily mean net job losses – on the contrary, BCG estimates a net increase of 350,000 jobs by 2025 – it does mean upheaval, both in the job market and in the political sphere.

On the job market front, Germany has a shrinking pool of skilled labour. The Association of German Chambers of Commerce and Industry (DIHK) consider this poses the biggest risk to their businesses. The government is acutely aware of the issue – its August 2017 progress report projects 700,000 fewer skilled workers in 2030 than in 2014. Moreover, with an ageing population, the demographics are currently not in Germany’s favour.

Resolving this issue will require big and difficult political changes. On the one hand, it means that more immigration, particularly of young skilled workers, will likely be necessary. Given the backlash to Merkel’s "welcome" policy at the height of the refugee crisis, an anti-immigration sentiment was stirred which was dormant before.

On the other hand, while new jobs will be available, this does not necessarily mean that from one day to another that those working, for example, in manufacturing, will be keen to move into the service industry or another occupation altogether. Nor does it mean they will want to, or even perhaps be capable of, reskilling to carry out new digital roles.

In the UK and the US, we recently witnessed how these labour market changes were one of the big factors associated with support for the protectionist and anti-immigration rhetoric of the Leave campaign for Brexit and Donald Trump for president.

In Germany, the regions most exposed to the effects of automation are in the industrial south and west – parts of the country so far spared from the worst of populism. The potential for populist support to expand at the national level should therefore worry observers. To its credit, the current government has already been thinking about it, as evidenced in the Work 4.0 White Paper.

However, policy choices in the next few years will be crucial for mitigating the future labour market and political shockwaves of automation. If politicians choose to merkeln (do nothing) on the issue, the populist backlash might hit Germany, too.

Claudia Chwalisz is a consultant at Populus and a fellow at the Crick Centre, University of Sheffield. She is the author of The People’s Verdict: Adding Informed Citizen Voices to Public Decision-making (2017) and The Populist Signal: Why Politics and Democracy Need to Change (2015). Her guide to the German election authored with Matthew Elliott can be downloaded here

© New Statesman 1913 - 2017ABOUT US

[FaxModem1]

PBS has done a video about this topic:

http://www.pbs.org/newshour/bb/brink-jobless-future/

Things are not looking good.

[FaxModem1]

CNBC

McDonald's hits all-time high as Wall Street cheers replacement of cashiers with kiosks
Cowen says McDonald's will upgrade 2,500 restaurants to its "Experience of the Future" technology by year-end, which includes digital ordering kiosks.
The firm raises its rating on McDonald's to outperform from market perform and price target for the shares to $180 from $142.
Same store sales estimate for 2018 raised to 3 percent from 2 percent.
Tae Kim | @firstadopter
Published 11:30 AM ET Tue, 20 June 2017 Updated 4:22 PM ET Thu, 22 June 2017
CNBC.com
PLAY VIDEO
McDonald's shares hit an all-time high on Tuesday as Wall Street expects sales to increase from new digital ordering kiosks that will replace cashiers in 2,500 restaurants.

Cowen raised its rating on McDonald's shares to outperform from market perform because of the technology upgrades, which are slated for the fast-food chain's restaurants this year.
McDonald's shares rallied 26 percent this year through Monday compared to the S&P 500's 10 percent return.
Andrew Charles from Cowen cited plans for the restaurant chain to roll out mobile ordering across 14,000 U.S. locations by the end of 2017. The technology upgrades, part of what McDonald's calls "Experience of the Future," includes digital ordering kiosks that will be offered in 2,500 restaurants by the end of the year and table delivery.

"MCD is cultivating a digital platform through mobile ordering and Experience of the Future (EOTF), an in-store technological overhaul most conspicuous through kiosk ordering and table delivery," Charles wrote in a note to clients Tuesday. "Our analysis suggests efforts should bear fruit in 2018 with a combined 130 bps [basis points] contribution to U.S. comps [comparable sales]."

He raised his 2018 U.S. same store sales growth estimate for the fast-food chain to 3 percent from 2 percent.

The analyst raised his price target for McDonald's to $180 from $142, representing 17.5 percent upside from Monday's close. He also raised his 2018 earnings-per-share forecast to $6.87 from $6.71 versus the Wall Street consensus of $6.83.
"MCD has done a great job launching popular innovations within the context of simplifying the menu, while introducing more effective value initiatives that have recently begun to improve the brand's value perceptions," he wrote.
A McDonald's spokeswoman sent the following statement in response to this story:
"Our CEO, Steve Easterbrook, has said on many occasions that self-order kiosks in McDonald's restaurants are not a labor replacement. They provide an opportunity to transition back-of-the-house positions to more customer service roles such as concierges and table service where they are able to truly engage with guests and enhance the dining experience."
— CNBC's Michael Bloom contributed to this story.

Question for the economists out there, how long can McDonald's stock continue to rise while offloading workers to machines? Would this be a short term gain or a long term one?

Discuss

[Gandalf]

Would this be a short term gain or a long term one?

Long term. By cutting the required amount of components between customer and burger, they lower expenses. These can just be taken as profits, used to lower prices, or even to be reinvested somewhere in the restaurant. Some can become smaller, bare bones restaurants, and some can become bigger and fancier in nature.

[FaxModem1]

MIT Technology Review

Andrew Ng Wants a New “New Deal” to Combat Job Automation

AI is coming for jobs, but one of the field’s masters has an idea that he thinks can help.

by Rachel Metz November 7, 2017
18

h
Andrew Ng would like to see a "new New Deal" that pays people displaced by technology to study, offering an incentive to learn new skills and reenter the workforce.
Recommended for You
IBM Raises the Bar with a 50-Qubit Quantum Computer
A Mind-Bending Cryptographic Trick Promises to Take Blockchains Mainstream
Could the AI Talent Shortage Be Eased By Tech Giants Learning to Share?
Quantum Computers Pose Imminent Threat to Bitcoin Security
Is Technology About to Decimate White-Collar Work?
Andrew Ng, formerly the head of AI for Chinese search giant Baidu and, before that, creator of Google’s deep-learning Brain project, knows as well as anyone that artificial intelligence is coming for plenty of jobs. And many of us don’t even know it.

Speaking at MIT Technology Review’s annual EmTech MIT conference in Cambridge, MA, on Tuesday, Ng said he’s visited call centers and spoken to workers, knowing that his teams of software engineers will then write software that will automate aspects of their work.

“There are many professions in the crosshairs of AI teams across the world,” he said.

Ng, who’s currently working on a startup called Deeplearning.ai that helps train people on deep-learning technology, has some ideas for helping those in jobs he thinks will be automated, from call-center workers to radiologists, truck drivers, and the like.

His suggestion is for an updated version of the New Deal—the Depression-era economic programs that invested in, among other things, getting unemployed Americans back to work—that pays displaced workers to learn new job skills.

Even if AI does not advance beyond today’s state of the art, Ng said, he believes it’s already capable of transforming all kinds of industries. Just supervised learning, the technique where a computer is given an input (such as a picture of a face) and learns to use that to predict an output (such as whether that face is you), is enough.

“Even now, if we stopped writing research papers, we have enough to transform the industry,” he said. Future waves of innovation will bring even more profound changes to job markets, and we will need a way to continue to adapt to them.

Would this be viable, or a stopgap measure?

[bilateralrope]

Would this be viable, or a stopgap measure?

A bit of both. In terms of fighting mass unemployment due to automation, it will only delay the problem.

At as means of keeping the population educated it has potential.

Google’s Machine Learning Software Can Create Code Better Than The Researchers Who Made It

Google’s AutoML system, which was introduced back in May, has recently produced machine-learning codes with higher efficiency than those made by the researchers themselves. Google developed AutoML to be an artificial intelligence that could help humans create other self-learning systems.

AutoML was developed because there are few experts with the knowledge to build highly complex AI systems. To meet the demand for experts, Google developed AutoML to create self-learning code, and in a way, clone itself.

Google reported in its official blog that AutoML can be trained and evaluated for quality on particular tasks. Feedback is generated to improve the proposals for the subsequent rounds. AutoML can run thousands of simulations to make appropriate changes, generate new architectures, and give recurring feedback.

Whats more exciting is that AutoML recently scored an 82% efficiency while sorting images based on their content. It was also given the task of locating multiple objects within an image. The code that AutoML wrote managed to score 43% while the best human-built program scored 39%.

AutoML is just our machines automatically generating machine learning models. Today, these are handcrafted by machine learning scientists and literally only a few thousands of scientists around the world can do this, design the number of layers, weight and connect the neurons appropriately. It’s very hard to do. We want to democratize this. We want to bring this to more people,” said Google CEO Sundar Pichai at the October 4 event.

It is noteworthy that instead of designing a neural network from the ground up, it is efficient to work on an existing AI system and modify it according to the task it has to perform. Humans will merely have to play a gatekeeping role as more of these intelligent systems start to take over our work and outperform us.

Do you think we are approaching singularity with such self-learning systems? Share your views in the comments.

[FaxModem1]

If that's the case, does this mean that even the technological fields are in danger, as the computer would be able to reprogram itself, or would that mean an increase of programmers, to check what the computer is doing to itself?

[K. A. Pital]

If that's the case, does this mean that even the technological fields are in danger, as the computer would be able to reprogram itself, or would that mean an increase of programmers, to check what the computer is doing to itself?

The technological fields are the ones most endangered by automation, actually. Purely manual, hard-to-automate tasks where heuristical approaches still rule the day, are likely to last way longer than the formalized programming workspaces. Only the informal, high-level goal- and objective-setting positions will remain occupied by humans with certainity. The rest of currently popular "coder" jobs will be wiped out.

[Ziggy Stardust]

The technological fields are the ones most endangered by automation, actually. Purely manual, hard-to-automate tasks where heuristical approaches still rule the day, are likely to last way longer than the formalized programming workspaces. Only the informal, high-level goal- and objective-setting positions will remain occupied by humans with certainity. The rest of currently popular "coder" jobs will be wiped out.

Indeed.

Just look back at the history of the word "computer" for an example. Up until only a few decades ago, complex computational tasks were essentially manual, with the work being brute forced by a team of human "computers" (that's literally the origin of the term, as 'one who computes'). You could have several dozen or even hundreds of computers working on various components of the calculation (this was done, for example, by NASA and in meterology, liquid dynamics, and other engineering fields). Starting in the late 1950s and 1960s, these teams of computers were replaced by early modern electrical computers (the famous room sized ones programmed using stacks of punch cards and the like). These computers still required a dozen or more people to operate and program, but it was a massive downsizing from the large teams of human computers that had been used up to that point. About a decade later (late 1960s and early 1970s), the first personal computers were being developed, which further downsized the requisite workspace.

[Simon_Jester]

If that's the case, does this mean that even the technological fields are in danger, as the computer would be able to reprogram itself, or would that mean an increase of programmers, to check what the computer is doing to itself?

The technological fields are the ones most endangered by automation, actually. Purely manual, hard-to-automate tasks where heuristical approaches still rule the day, are likely to last way longer than the formalized programming workspaces. Only the informal, high-level goal- and objective-setting positions will remain occupied by humans with certainity. The rest of currently popular "coder" jobs will be wiped out.

One of the biggest hard-to-automate tasks involves machine vision; the very class of neural networks referenced in the article tends to reduce the degree to which that's an obstacle.

If you can fully solve machine vision (to the point where a robot can, say, detect a warning sign as a warning sign, read its text, and respond appropriately to the warning)... You've greatly increased the range of manual tasks that can be automated.

Honestly, there are an arbitrary number of stories we can make up about why this or that field is more or less vulnerable to automation than other fields. I don't think we can really predict it anymore, except where there is already an existing trend happening as we speak (i.e. self-driving cars, which have not YET displaced any drivers, but will probably displace vast numbers of them within the next generation).

[bilateralrope]

If you can fully solve machine vision (to the point where a robot can, say, detect a warning sign as a warning sign, read its text, and respond appropriately to the warning)... You've greatly increased the range of manual tasks that can be automated.

That is a problem that self driving cars have. Have a look at the pictures to see just how bad the problem can be.

[Simon_Jester]

Yeah, I kind of figured that was an issue. That does illustrate, though, that machine vision is still a "half-solved" problem. Probably unsurprising; vision processing is one of the cognitive tasks that evolution has been optimizing animal brains for longest of all. Our visual processing is, on the whole, very hard to fool and very hard to hack under realistic environmental conditions.* It would be a big surprise if it were anything other than one of the last human tasks we figure out how to automate.**
____________________

*There are optical illusions, sure, but the thing is, our brain's processing tends to parse them as "huh, what a freaky picture," and move on with life, which is exactly what a robust system should do when confronted with an anomalous input that doesn't actually make any difference.

**At least, possibly excluding tasks that require sentience, such as social interactions that make use of theory of mind and so on.

[Wild Zontargs]

Take it one step further, and you can generate "adversarial images" which look like one thing to humans, but something else entirely to AIs:

Shotgun shell: Google's AI thinks this turtle is a rifle

If it is the shape of a turtle, the size of a turtle, and has the patterning of a turtle, it’s probably a turtle. So when artificial intelligence confidently declares it’s a gun instead, something’s gone wrong.

But that’s what researchers from MIT’s Labsix managed to trick Google’s object recognition AI into thinking, they revealed in a paper published this week.

The team built on a concept known as an “adversarial image”. That’s a picture created from the ground-up to fool an AI into classifying it as something completely different from what it shows: for instance, a picture of a tabby cat recognised with 99% certainty as a bowl of guacamole.

Such tricks work by carefully adding visual noise to the image so that the bundle of signifiers an AI uses to recognise its contents get confused, while a human doesn’t notice any difference.

But while there’s a lot of theoretical work demonstrating the attacks are possible, physical demonstrations of the same technique are thin on the ground. Often, simply rotating the image, messing with the colour balance, or cropping it slightly, can be enough to ruin the trick.

The MIT researchers have pushed the idea further than ever before, by manipulating not a simple 2D image, but the surface texture of a 3D-printed turtle. The resulting shell pattern looks trippy, but still completely recognisable as a turtle – unless you are Google’s public object detection AI, in which case you are 90% certain it’s a rifle.

The obvious next question is "can I make a hypothetical police/security AI think my rifle is a turtle?"

[K. A. Pital]

One of the biggest hard-to-automate tasks involves machine vision; the very class of neural networks referenced in the article tends to reduce the degree to which that's an obstacle.

If you can fully solve machine vision (to the point where a robot can, say, detect a warning sign as a warning sign, read its text, and respond appropriately to the warning)... You've greatly increased the range of manual tasks that can be automated.

Honestly, there are an arbitrary number of stories we can make up about why this or that field is more or less vulnerable to automation than other fields. I don't think we can really predict it anymore, except where there is already an existing trend happening as we speak (i.e. self-driving cars, which have not YET displaced any drivers, but will probably displace vast numbers of them within the next generation).

Even if you can fully solve machine vision, you can't fully solve the problem of machine response vs. human response. But in IT, human response and human approach to tasks and goals are becoming less and less relevant. Human response remains partially relevant in driving, in healthcare, in many instances of human-human interaction. But IT and programming is centered around human-machine interaction, so it is by definition an area that sees the greatest labour-saving efficiencies from advances in itself.

[Simon_Jester]

The converse of this is that we've already automated a huge number of manual jobs or jobs that traditionally required human interaction. This already happened, so it's unlikely that we've reached the endpoint of the process.

'Travel agent' used to be a more common job; think of all the human interaction involved! Until it turned out people could book their own airplanes and hotels on the Internet and do reasonably well. 'Shoemaker' used to be a more common job, because it was performed by individual cobblers making individually fitted shoes... until it turned out to be more efficient to have factories mass-produce shoes and let everyone find the size that suited them best.

I'm not saying you don't have a valid argument for why IT and programming can or will be automated. I'm saying that you don't have an argument for why that field will automate faster than everything else.

[bilateralrope]

The obvious next question is "can I make a hypothetical police/security AI think my rifle is a turtle?"

To the robots visual camera, probably. But there are two possible solutions to that problem:
- Better image recognition.
- Other sensors. If the robot sees what it thinks is a turtle, but its metal detector detects too much metal, you haven't fooled it.

Well, assuming that the robot doesn't commit suicide.

[K. A. Pital]

I'm not saying you don't have a valid argument for why IT and programming can or will be automated. I'm saying that you don't have an argument for why that field will automate faster than everything else.

I'm not saying that it will; but it tends to follow a logical assumption that a sector which is new and where automation has started to reduce labour inputs will keep undergoing the same process at a faster rate than old sectors where most automation gains have been picked by now already...

[Simon_Jester]

The obvious next question is "can I make a hypothetical police/security AI think my rifle is a turtle?"

To the robots visual camera, probably. But there are two possible solutions to that problem:
- Better image recognition.
- Other sensors. If the robot sees what it thinks is a turtle, but its metal detector detects too much metal, you haven't fooled it.

Maybe you have, maybe you haven't. The robot may or may not be programmed to aggressively go after people it sees carrying steel turtles around.

I'm not saying that it will; but it tends to follow a logical assumption that a sector which is new and where automation has started to reduce labour inputs will keep undergoing the same process at a faster rate than old sectors where most automation gains have been picked by now already...

I think that's applying a level of 'logic,' or rather abstraction, that loses sight of too many important details.

As different technologies become available, different fields become automated easily and rapidly. Ditch-digging, for example, became automated with the rise of mechanical excavators, but virtually no computer technology was required. Clerks and accountants found a lot of their work becoming automated very, very quickly with the rise of personal computers in the late 20th century, but no artificial intelligence was required to automate most of that sector.

Almost no jobs have yet been lost by human drivers to robot cars... but some time in the next decade or three, we will hit an inflection point at which robot cars are seen as a safe, practical alternative to human drivers. At which point very, very suddenly a lot of automobile jobs may be lost very, very rapidly.

[The Romulan Republic]

Long-term, I think that we're going to see a world where their are fewer and fewer jobs available, unless something very unexpected and probably catastrophic happens.

In that case, we can respond in three ways:

1. Double-down on blaming the poor for being poor, until eventually people turn to extremist ideologies and armed revolt out of desperation.

2. Double-down on scapegoating minorities, which will likely culminate in fascism and genocide. It won't solve the underlying problem though- just let those in power defer the consequences for a time by putting them on the backs of innocents.

3. Basic income. Create a society where most people don't have to work to live decent lives. This will fuel resentment from those who see the poor as lazy leeches living off their dime*, but a combination of everyone getting to be a "leech", eliminating red tape by rendering certain social programs redundant, and shear necessity may make it more palatable.

I honestly consider Basic Income the single most pressing economic issue of our times, because I am of the opinion that it is very likely a choice for our society between that and genocidal despotism in the fairly near future (ie, the next few decades). This informs my vote, and is one of the main reasons why I backed the British Columbia Green Party last election up here.

*By the way, I don't buy the idea that if people don't have to work to live, they'll all become lazy parasites. Take it from someone who's spent a lot of his adulthood unemployed- doing nothing is boring. When I don't have a job, I volunteer, partly because I'd probably go mad if I didn't. Most people, I think, like to work- not necessarily at something that makes a lot of money (like me and theatre, for instance), but just because something doesn't usually turn a high profit, it doesn't mean that it doesn't contribute to society. Providing a safety net for everyone might actually encourage more people to volunteer, or to try to start a business of their own, because they don't have to worry about getting a pay cheque every week to live.

Then again, I also don't feel that the right to live (which requires, at minimum, food, basic medical care, and shelter unless you live somewhere with a very temperate climate) is a privilege you have to earn.

Yes, I basically want to turn our economy into that of the United Federation of Planets. Sue me.

[K. A. Pital]

The problem with basic income is that under capitalism with prices, markets and inflation, it would very likely cause a loss of purchasing power and very soon people on it would have just barely enough to maintain adequate caloric intake. I mean, there is a problem of working poor, and if the "basic income" is about as much as low-end wages, it certainly won't solve these issues.

It is kind of like working-till-death retirees who have to keep working because the erosion of pensions' purchasing power leaves them with few other options. This actually happened in some countries undergoing severe inflation.

[Zaune]

The problem with basic income is that under capitalism with prices, markets and inflation, it would very likely cause a loss of purchasing power and very soon people on it would have just barely enough to maintain adequate caloric intake. I mean, there is a problem of working poor, and if the "basic income" is about as much as low-end wages, it certainly won't solve these issues.

It is kind of like working-till-death retirees who have to keep working because the erosion of pensions' purchasing power leaves them with few other options. This actually happened in some countries undergoing severe inflation.

It won't, but at least a UBI that guaranteed adequate food and shelter would be a start. At least with a UBI you don't have to worry about arbitrary conditions like "make a penny over £100 a week and we'll take away your £180 a week disability allowance" so you can feel free to top it off any way you can think of.

[Starglider]

If that's the case, does this mean that even the technological fields are in danger, as the computer would be able to reprogram itself, or would that mean an increase of programmers, to check what the computer is doing to itself?

AI research advances tend to get misreported/misunderstood almost as much as quantum physics. AutoML is a useful development in deep neural network engineering, but it doesn't write code, and to be clear a lot of people already had this idea, they just didn't have enough data, compute and use cases to make it really useful. What it does is automate the rather convoluted and tedious NN parameter tuning process, which breathless press releases about the power of deep learning have always glossed over. In other words, it makes ANNs work more like their marketing claims of automatically generating a mapping function for any classification or feedback/control task. The actual engine code that implements the ANN, and the substantial amount of auxiluary (conventional) code required to make any useful ML-powered software system, is not generated or modified by AutoML. This is a substantially harder but also less pressing problem, because NN engine code can be written by a few experts and reused by everyone, and the auxiluary code can be cranked out by standard software engineers.

[The Romulan Republic]

The problem with basic income is that under capitalism with prices, markets and inflation, it would very likely cause a loss of purchasing power and very soon people on it would have just barely enough to maintain adequate caloric intake. I mean, there is a problem of working poor, and if the "basic income" is about as much as low-end wages, it certainly won't solve these issues.

For whatever its worth, I believe that any legislation establishing basic income should tie it to inflation, so that the amount one receives is automatically adjusted to match inflation.

It also needs to be done in conjunction with seriously addressing the lack of affordable housing in a lot of places.

It is kind of like working-till-death retirees who have to keep working because the erosion of pensions' purchasing power leaves them with few other options. This actually happened in some countries undergoing severe inflation.

Yeah, no law or policy is likely to fair very well forever if it does not keep pace with the demands of the times. Doesn't mean that the underlying idea is a bad one. Just that their are no easy, permanent fixes, and that we have to take these issues into account.

The problem with basic income is that under capitalism with prices, markets and inflation, it would very likely cause a loss of purchasing power and very soon people on it would have just barely enough to maintain adequate caloric intake. I mean, there is a problem of working poor, and if the "basic income" is about as much as low-end wages, it certainly won't solve these issues.

It is kind of like working-till-death retirees who have to keep working because the erosion of pensions' purchasing power leaves them with few other options. This actually happened in some countries undergoing severe inflation.

It won't, but at least a UBI that guaranteed adequate food and shelter would be a start. At least with a UBI you don't have to worry about arbitrary conditions like "make a penny over £100 a week and we'll take away your £180 a week disability allowance" so you can feel free to top it off any way you can think of.

Also this, though I think his concern is that it might end up not being enough to cover "adequate food and shelter".

One would hope that under basic income, most people would still do some form of work, either out of personal enjoyment or because they want more than simply "basic" income. But the basic income would provide a minimum standard, a safety net, beneath which no one would be permitted to fall.

[K. A. Pital]

The problem with basic income is that under capitalism with prices, markets and inflation, it would very likely cause a loss of purchasing power and very soon people on it would have just barely enough to maintain adequate caloric intake. I mean, there is a problem of working poor, and if the "basic income" is about as much as low-end wages, it certainly won't solve these issues.

For whatever its worth, I believe that any legislation establishing basic income should tie it to inflation, so that the amount one receives is automatically adjusted to match inflation.

Since typical inflation measurements does not adequately reflect the erosion of wage/income purchasing power (cars and especially education, healthcare and houses have become much more costly nowadays), I doubt that it would stop the described scenario. Alleviate? Perhaps. But not stop.

It also needs to be done in conjunction with seriously addressing the lack of affordable housing in a lot of places.

That's just capitalism at work. Are you proposing government-constructed housing to crash prices in the housing market? If not, then I fear status quo will remain. But if you do, think about what this crash will do to the other elements of capitalist economy strongly tied to it - finance & banking with enormous credit volumes, construction companies, etc.