Returning to work and AI confusion on The Future, This Week

This week: will Silicon Valley finally move online? And AI is confused by our weird behaviour. Sandra Peter (Sydney Business Insights) and Kai Riemer (Digital Disruption Research Group) meet once a week to put their own spin on news that is impacting the future of business in The Future, This Week.

The stories this week

02:18 – Tech companies considering letting employees work from home forever

16:15 – AI gets confused by our weird behaviour during the pandemic

23:11 – Facebook to embrace hiring of remote workers

Other stories we bring up

Even the pandemic can’t kill the open-plan office

Over 25% of tech sector wants permanent work from home

Our previous conversation around office space on Corona Business Insights

Our previous conversation around productivity and remote work on Corona Business Insights

The end of the office as we know it

What will tomorrow’s workplace bring

Facebook predicts that 50% of the company’s employees could be working remotely within the next 5-10 years.

Minimum wage for H-1B visa holders could reach $250,000 a year


You can subscribe to this podcast on iTunesSpotifySoundcloudStitcherLibsyn, YouTube or wherever you get your podcasts. You can follow us online on FlipboardTwitter, or sbi.sydney.edu.au.

Our theme music was composed and played by Linsey Pollak.

Send us your news ideas to sbi@sydney.edu.au.

Disclaimer We'd like to advise that the following program may contain real news, occasional philosophy and ideas that may offend some listeners.

Intro This is The Future, This Week on Sydney Business Insights. I'm Sandra Peter, and I'm Kai Riemer. Every week we get together and look at the news of the week. We discuss technology, the future of business, the weird and the wonderful and things that change the world. Okay, let's start. Let's start!

Sandra Today on The Future, This Week: will Silicon Valley finally move online? And AI is confused by our weird behaviour. I'm Sandra Peter, I'm the Director of Sydney Business Insights.

Kai I'm Kai Riemer, professor at the Business School and Leader of the Digital Disruption Research Group.

Sandra So what should we talk about today?

Kai Um, well there was something about AI camera glasses, so glasses with a little camera, and it would then read any text that is in front of you. There's also rumours that Apple is now in the final stages of working on its augmented reality glasses. But that's not enough to actually carry a story, something to keep an eye on.

Sandra There are some updates on stories we spoke about previously. Uber has just laid off another 3000 people. There are some updates on food delivery apps and arbitrage using food delivery apps.

Kai But really the story that's been dominating the news is how and when are we going back to the office?

Sandra But we've had an episode on what the office might look like once we get back on Corona Business Insights and we'll put the link in the shownotes. There was a set of very surprising stories around the fact that some of the people with the coolest offices and the nicest perks do not really want to go back.

Kai So Silicon Valley and going back to the office, let's do that.

Sandra And there is another thing that we should really talk about today, because all our weird behaviour during the COVID-19 pandemic is really messing with the models that many of the companies that we use to deliver our products and our services have been employing for years now. And it's forcing us to step in and do something about it.

Kai Let's do this. So, Sandra, what happened in the future this week?

Sandra Our first article's from Wired, but there have been a whole range of articles on the topic, and it's on the fact that some of the employees of the big tech companies do not want to go back to work. So Silicon Valley is considering letting many of their employees give up the commute forever.

Kai So Twitter, for example, came out and announced that people can, from September possibly, go back to the office, but they don't have to. Not only do they not have to come back this year, but they can work from home indefinitely. This is a big change for Twitter or indeed for any of these Silicon Valley companies in this many who are thinking about following in Twitter's footsteps because they've put so much effort and energy into building these great campuses, these theme park offices.

Sandra So indeed, both Twitter and Square have said that employees can work from home forever. Google and Facebook have told their employees that they can work from home until the end of this year, and all up CNBC reported about 27 percent of the tech sector workers say they prefer remote work or working from home.

Kai And Facebook has announced that it will reopen offices starting in July, but only at 25 percent capacity.

Sandra And for now, Apple seems one of the few companies that is actually asking people to slowly come back to work. But then Apple has seldom allowed people to work from home due to its culture of secrecy and its emphasis on people being present at work and having in-person meetings.

Kai Yeah, and let's also not forget that Apple, unlike Google, Facebook or Twitter, which are mostly software companies, Apple is a hardware company and it is really not that easy to prototype, to design, to test hardware in your living room. So there's a certain infrastructure that goes into much of what Apple does. On top of that, as you said, of course, the secrecy. There have been instances in the past where people used to prototype Apple devices like a new unreleased iPhone. Famously, someone forgot this in a pub, and a whole saga ensued from that. So Apple has always been quite secretive, and so they want their people close on campus. But they're very much the exception, it seems, in this current climate.

Sandra And interestingly, what we've seen over the last decade or so, was many of these companies going back and forth from creating the tools and asking people to embrace remote working, to asking people to come back to the office, with famously seen in 2013, Marissa Mayer from Yahoo! banning telecommuting and saying that speed and quality of work are sacrificed when people do not come into the office. IBM used to be a remote work proponent, asking people to consider flexible working arrangements, and three years ago asked their teams to come back to work in the office, work 'shoulder to shoulder', is now again embracing working remotely as they've done over the past couple of months.

Kai Well it has always been somewhat ironic that the very companies who would build the infrastructure for distributed working, digital working, remote working, would put so much emphasis on having their teams co-located, and building these great offices with free food, catering, Google has done a lot of work around how best to design the healthiest, most productive face-to-face office environments. I had the chance to visit the offices in San Francisco of the then startup Yammer, which is now, of course a Microsoft product. And even though all of their work would happen online and much of their work in their own tool, they put so much emphasis on communal kitchens and being co-located and having team boards, so this whole idea of the physically co-located, agile work practices. So a real shift for many of these companies to now abandon this and going to remote working.

Sandra And you're right, apple's headquarters, Google, Facebook, they have tremendous facilities with restaurants, with places to relax or play games with every kind of supply you might want to take advantage of, from language classes to getting massages to being able to bring your dog to work or eat food on any kind of diet. They're tremendous places to work and they are a big culture and part of what attracts people to work at these companies. But that's all up for grabs now.

Kai But people who in these companies, they give up a lot by going to maybe work in their tiny apartments, because who needs a great apartment when you have all these great facilities and many of your friends at work and you can hang out there.

Sandra But there's also the question of who can afford big apartments. Let's remember, many of these companies are located in Silicon Valley and California, San Francisco, Palo Alto, it's a very, very expensive place to live. So many of these people share houses, very small apartments and spend most of their time at work.

Kai And let's face it, we've both been there. The traffic in the valley is a bit shit, right? So much of the time for many of these people is, in fact, spent commuting, sitting in traffic. That, of course, then makes it a whole different equation if you can spend two or three hours less on the road staying at home. That might make it easier to forego some of these perks.

Sandra But clearly a large proportion of these people, at least a quarter of them, do say they want to work from home, which opens up a whole new set of questions that don't have anything to do with offices, but rather with salaries.

Kai If we think a little bit further to where things might lead once we actually open this box, there's a bunch of ripple effects that might ensue in the future. So in the article, people are being quoted saying, if I'm working from home anyway, why would I then pay Bay Area rent, which is exorbitant? You know, I might as well move away. I don't have to live in San Francisco or Palo Alto. I can move interstate and have a much cheaper lifestyle and still do my work online.

Sandra But that, of course, also means that the pool of employees for companies like Google and Facebook might now extend to people who do not want to move to the valley, who might live elsewhere in the US or indeed elsewhere in the world. So the competition for those jobs might become quite different. Also, the wages paid might change, because indeed I do not have to motivate people to move to San Francisco and live there.

Kai Yeah, a lot of these salaries are a premium because living in Silicon Valley and surrounds is so expensive. So when people move away, maybe companies would say we don't have to pay those wages because the cost structure has now changed.

Sandra So it's interesting to consider what a decentralised workforce would look like in the tech industry, or indeed a decentralised tech industry, that might now attract people from other industries and people who would have not considered this because it required a move to California.

Kai And there's an article in CNBC which raises the prospect that we might actually see a brain drain from other industries. So once Silicon Valley starts casting a wider net and drawing from a wider pool of potential workers they might, and this is the fear actually snatch up talent that is not located or not prepared to move to Silicon Valley because now they can work from anywhere.

Sandra This would, of course, depend also on where salaries would go. If we do see a huge decrease in average salaries because some of the people are not there, people might not be so enticed to move. But on the other hand, if salaries still stay fairly high, it might attract a lot of top talent from other companies back to the tech industry.

Kai Yeah, but then let's take this one step further even. What does Silicon Valley actually mean? What does it stand for? So at this point in time, moving to Silicon Valley means I'm in the close proximity and I run into people in coffee shops. I can mix and mingle with a lot of like-minded people and I'm exposed to ideas, but I also have ready access to funding. And there is a location aspect to this. There's also an identity aspect to it moving there. So it attracts people. It has a certain magnetic force to it. Once that disperses, why would people want to then work for Silicon Valley firms when they're not really Silicon Valley firms anymore? What would it do to the brand Silicon Valley, so to speak, going forward?

Sandra Let's remember, it's only about 25 percent of employees who are saying they would like to work from home or work virtually, and Silicon Valley still became what it is today because of that co-location of all these innovative companies, but also offer whole ecosystem of universities, Stanford, Berkeley and innovation centres that many of the large corporations have there, regardless of the industry that they're in.

Kai And that's the point, isn't it, because the current lot of companies, the big brands, Google, Twitter, Apple, we know that they are Silicon Valley companies. But what happens to all these start-ups who want to be associated with the tech industry in Silicon Valley? There is a good case to be made why it makes sense to be there to have access to that localised talent pool and to actually have your offices, or at least some presence in Silicon Valley, even though it might not mean that everyone is in the office, that it is virtualised to a greater extent. But I, for one, cannot see how the location Silicon Valley will be greatly diminished through this.

Sandra But at least in the short-term, what it looks like will change quite dramatically. So we mentioned before that Facebook is coming back with 25 percent of its workforce. But most workplaces are looking at reducing capacity by 50 to 60 percent, especially now when we're still just coming out of lockdown. These offices will look very different, and the interaction in them will look very different, at only 25 to 50 percent capacity.

Kai And that's not just a Silicon Valley phenomenon. Of course, that is true for many of the offices around the world. In Sydney here, we're just at the cusp of opening things up and some companies are moving some of their staff back to the office. But that raises new questions. We've discussed previously on Corona Business Insights that the office itself might change that it feels different, there's more cleaning, the office furniture might change. But it also raises questions for how to actually organise work and why people might want to go back to the office or not.

Sandra And also raises significant challenges when only part of the workforce is back. In Silicon Valley, as with companies in Sydney, when only part of your workforce is at the office and the rest are working remotely, organising work, organising collaboration, ensuring that everybody feels part of the same team, becomes a lot more complicated.

Kai And let's not forget that many companies have in place business continuity plans to reduce the risk of exposure. They split up divisions or they split up teams to make sure that not every knowledge bear or not every senior person is in the office at once, so that if there is a virus spread in the office, it doesn't take out every critical person in a division. So the people who are actually closely working together might not actually be in the office at the same time, which raises questions, then why would half the team go to the office if they have to be on Zoom anyway? Why would I have to face what is now a much more cumbersome commute? Why would I have to queue at the elevator that only takes one or two people at the time. If I can't just stay at home.

Sandra So what this means is that going back to the office is not just about how we redesign the physical space, how we reconfigure workspaces or what technologies we use. But it's also about how we rethink roles, how we rethink our organisational cultures and the tasks and the activities that we engage in.

Kai And maybe some companies might want to be a bit more flexible or experiment a little bit more to rethink the idea of the office off the back of this. The question sometimes is why would someone go back to the office? For many people, it might be because working at home is just not productive, there's too many distractions, there might be lack of space, the internet at home might not be great, but going back to the main office might equally be unappealing because of the commute or the wait times at the elevator. So maybe providing some offsite office space, a decentralised office network in co-working spaces might be one way to provide little islands pockets of people working together without having to make them move back to one big office as we were used to in the past.

Sandra And indeed, this is probably the way forward. Whilst the pandemic was one big, unprecedented experiment, organisations might want to run smaller, rapid experiments in workplace redesign in decentralising the office and then figuring out what works best. Given the nature of the work, one size will probably not fit all.

Kai Speaking of things that work or don't work. Our second story has to do with AI and how AI models struggle with the weird behaviour that many of us all of a sudden display as life has changed so dramatically during the pandemic.

Sandra Our second story comes from the MIT Tech Review, and it reports on the fact that once COVID-19 hit, we really changed what we bought online what cues we used to make our decisions. So for instance, people flocked to hand sanitizers and toilet paper and masks, decided that they're gonna buy things not based on their price, but on how quickly they would be delivered.

Kai In Amazon, when previously phone cases, phone chargers, electronics, Lego even, dominated the top 10 in what people buy, this all got replaced almost overnight and the game changed quite dramatically.

Sandra And this caused some significant problems to the machine learning algorithms that run behind the scenes, not only to give suggestions as to what we would buy next, but that decide what products or services gap advertising money, how inventory is managed, how stock is moving around the country from warehouse to warehouse or being dispatched to people. And all these algorithms are trained on average behaviour. They are built on data that these companies have collected over the years, and that exhibits specific trends. And of course all of these algorithms to accommodate change, change in preferences, change in behaviour, but not at the scale that we have seen with COVID-19.

Kai So I think it's worth briefly pointing out how these algorithms work, right? Unlike traditional computing where someone sits down and programs a set of instructions into an algorithm in a high, models are deduced from large amounts of data, and that data describes our collective past behaviours depending on context, what things we buy or what things we search for in a search engine, the things that get the most clicks in advertising. And so deriving predictions about what people will do next from past data works really, really well as a way of automating things when we're in a reasonably stable environment where the past is a good predictor of future behaviour. When a shock hits like COVID-19 and the automation of warehouses, of advertising, of search engines, of restocking, of what things will sell, goes out the window and people really have to step in.

Sandra And of course, we've discussed this many times before and previously we've pointed out just how difficult it is to actually observe when these algorithms get things wrong, when they do encounter anomalies, for instance, when they give out loans or decide who's suitable for a job ad, that they wouldn't be good at accommodating things that do not fit their patterns. Well, right now, it's one of the few times when the limitations of these algorithms are really visible.

Kai Absolutely. So this is the big one, right? So we've said many times that they work really well and the stable environments. Now, the situation is so volatile that these models in many instances become useless. And then it's back to people who have to tweak those algorithms or override what these algorithms would recommend. So it's really an exercise of very ad hoc bug fixing and handling things in quite a crisis mode. So Amazon for example had to readjust, they had so much strain in their warehouses that they tweaked some of their algorithms to areas that were not essential products, they would recommend products that were not fulfilled through their own warehouses to reduce the stress on the system, because wait times for products were blowing out to multiple days or weeks, when in fact they usually want to promise overnight delivery. So a real crisis not just in the moving of the physical products, but also in tweaking the algorithms who all of a sudden are flying blind.

Sandra And there was another thing that the article mentions but that this pandemic also helped expose, which is how interdependent the algorithms that different companies use are. For instance, the fact that Amazon makes certain tweaks to its algorithm would then have knock on effects on algorithms that sellers to use to decide how much they want to spend on ads for their products and services. So every time a web page with ads loads, there is an auction where bidders bid automatically that decides the price of the ads that go on that page. And then there is also the additional layer of uncertainty as to what makes people buy in these times. So as new data trickles in, we see that, for instance, the decider might no longer be the price, but rather the speed of delivery for a product and service. So then it's not just about Amazon adjusting their algorithms, but all the other companies that are part of this ecosystem, they would all need to adjust their algorithms to reflect that.

Kai So data scientists in many of these companies are really in crisis mode. Let's not forget that these algorithms can't just be retrained on the fly because they rely on large amounts of reliable past data as we are in uncharted territory, as we are in volatile times. There is simply no data that predicts what people will do next. Which also means that automation of this particular kind, automation of the predictive kind from large amounts of data is not as useful during these disruptive shifts that we are seeing at the moment.

Kai And maybe that's one of the reasons why seemingly the conversation around the future of work that used to be dominated by automation and AI has now shifted back to people going back to work and people returning to the office.

Kai Yeah, I think we see that the future of work conversation has been completely shifted to new ways of working, digital work, remote work, flexible work. And that's understandable because it's the mode of work that we're coping with right now. But another reason really is that this data-based automation doesn't cope very well with volatility and rapidly changing times. So it's not the tool that helps right now, what we need at this point is humans sense-making, is humans stepping in, making those decisions, getting us through this crisis. And then I'm sure the conversation will shift back to AI and automation in the not too distant future.

Sandra And this is where we were going to leave you and we were going to finish up the podcast here, and we actually did. And while Megan was editing this podcast, breaking news:

Kai Overnight, Mark Zuckerberg announced that Facebook will aggressively embrace not only remote working, but hiring of remote workers who are not necessarily based in Silicon Valley, and isn't that what we've just discussed?

Sandra So we're adding a third story to today's podcast, and this one was really everywhere. And we're picking it up from The Guardian, and the story was titled "Facebook expects half of its employees to work remotely over the next five to 10 years". And it discusses Mark Zuckerberg's announcement to Facebook employees that they will be a forward leaning company on remote work and they will embrace this at scale, not only with regard to their current staff, who will have an opportunity under certain circumstances to choose to work remotely, but also with regard to hiring new people.

Kai And so, interestingly, there were some details which go to exactly what we've just discussed earlier on the podcast. One around salaries and the other around hiring from a wider talent pool.

Sandra Facebook has already announced that its employees do not need to come back until 2021, but they have now revised that to say that if employees want to work from a different location after the first of January next year, then under certain circumstances they will be allowed to work remotely and that they might have their compensation adjusted based on where they work from. So should these employees work outside of Silicon Valley, they will have their salaries adjusted to their location at that point.

Kai And he was being quite clear about it, saying that they will check up on employees, he said there will be severe ramifications for people who are not honest about this. So exactly what we discussed earlier, the fact that if you don't live in Silicon Valley in some of these really high-priced areas, then the temptation for companies will be to actually pay you less because you don't have to put up with the high rental prices, for example.

Sandra But this would also allow, in theory, Facebook to tap into a wider pool of talent, these would be people who want to work at these large tech companies but do not wish to live in big cities, or not even in the US. And that theoretically would also allow Facebook not only to attract talent, but also to potentially diversify its workforce.

Kai This, of course, is an interesting point, given that we've discussed so often the existence of a Silicon Valley bubble or the uniformity of people who live and work in Silicon Valley, and that in developing AI, for example, broader viewpoints are often not included because of the homogenous nature of the kind of people who build these algorithms. So maybe there's a point there. But also, it's not just a wider talent pool, right? There are salary implications. It might just be cheaper to hire people who are not living in Silicon Valley on these high salary packages, or indeed people who are not even living in the US.

Sandra And this comes at an interesting time, when the Trump administration is considering revising the minimum wage that is paid to people who are on an H1B visa. This would be foreign-born scientists, engineers, software developers who are considered very highly skilled and who are employed from around the world and end up working for US companies, maybe even assisting in this case with the country's economic recovery after COVID-19. And it's interesting to see that these revised estimates would put a minimum wage at between $150000 to 250000 a year. Currently, software developers on average make less than that. So this would put additional costs on companies like Facebook trying to hire talent from overseas.

Kai So as these companies prepare to work remotely, there might not be a need to get visas for these people and bring them into the country. There might be ways to bring them into a more permanent staff arrangement without them living in the US, going beyond labour for hire through gigwork platforms. But the more immediate announcement was that Facebook plans to hire up to 10000 engineers in three further hubs in the US, Atlanta, Dallas and Denver, where these people would be remote workers, but they would set up certain spaces where these people can meet not necessarily office, Zuckerberg says, but some kind of physical space for these engineers to occasionally work together, have meetings and discuss their work.

Sandra And given that in a recent survey of Facebook employees, 40 percent of them said they were extremely, very or somewhat interested in full-time remote work. And out of those, 75 percent said they were pretty confident they would move to different city if they were allowed to permanently work remotely, this might actually significantly change what Silicon Valley looks like in the future.

Kai And it might have implications for the tech industry or even other industries worldwide. Given that many companies look to the tech giants in Silicon Valley for ideas and they often do things first, and then these things are embraced by companies in other places. So it might well have implications for how Australian companies organise their labour force. And that is really all we have time for today.

Sandra More on the future of work both on Corona Business Insights and The Future, This Week.

Kai See you soon.

Sandra On The Future...

Kai Next week.

Sandra This week?

Kai Yes, but next week.

Sandra On The Future, This Week. Next week. Thanks for listening.

Kai Thanks for listening.

Outro This was The Future, This Week, made possible by the Sydney Business Insights team and members of the Digital Disruption Research Group. And every week right here with us, our sound editor Megan Wedge, who makes us sound good and keeps us honest. Our theme music was composed and played live on a set of garden hoses by Linsey Pollak. You can subscribe to this podcast on iTunes, Stitcher, Spotify, YouTube, SoundCloud or wherever you get your podcasts. You can follow us online on Flipboard, Twitter or sbi.sydney.edu.au. If you have any news that you want us to discuss, please send them to sbi@sydney.edu.au.

Related content

Processing...
Thank you! Your subscription has been confirmed. You'll hear from us soon.
ErrorHere