Will The Next Job Impacted By Automation Be App Development? (forbes.com) 149
Leading CIOs, CTOs and technology executives on the "Forbes Technology Council" just made some predictions for the future:
Now that the business world has seen the power of automation, the question has become, "What's next?" The members of Forbes Technology Council are constantly looking out for new tech trends, and they believe the next jobs to be impacted by automation might not be the ones people expect...
#1. Reminders, Notifications And Reporting
Christy Johnson, AchieveIt: I think as workflow technology expands, any kind of oversight-related job will be delegated to the bots. No human will be taking the time to manually build reports, see who they're missing data from and send those employees a reminder email/plea for a status update. The tech is already around, but I think it still has a long way to go to reach human-level logic and function....
#3. App Development
Katherine Kostereva, Creatio (formerly bpm'online): In the next five years, everyone will become a developer thanks to low-code/no-code technology. It allows users to build apps and processes in a visual integrated development environment with drag and drop features. Hand-coding isn't likely to become obsolete in five years, but we are moving towards a far future where little to no coding is involved in development.
#1. Reminders, Notifications And Reporting
Christy Johnson, AchieveIt: I think as workflow technology expands, any kind of oversight-related job will be delegated to the bots. No human will be taking the time to manually build reports, see who they're missing data from and send those employees a reminder email/plea for a status update. The tech is already around, but I think it still has a long way to go to reach human-level logic and function....
#3. App Development
Katherine Kostereva, Creatio (formerly bpm'online): In the next five years, everyone will become a developer thanks to low-code/no-code technology. It allows users to build apps and processes in a visual integrated development environment with drag and drop features. Hand-coding isn't likely to become obsolete in five years, but we are moving towards a far future where little to no coding is involved in development.
The Infinite Number of Monkeys (Score:4, Insightful)
With enough computing power AI will be able to write and test every algorithm possible. Once a company copyright's every iteration, no one will be a coder.
--
Automation applied to an inefficient operation will magnify the inefficiency. - Bill Gates
Comment removed (Score:5, Insightful)
Re: (Score:2)
Nicely said.
Re: (Score:2, Interesting)
Nice all around then, but once even close to everything is copyrighted, nothing is copyrighted.
Once you cannot speak three words or type two commands without infringing on protected pseudo-intellectual property, it all breaks down. No copyright protection is better than too much.
Re: (Score:2)
Re: The Infinite Number of Monkeys (Score:2)
Re: (Score:2)
HHGTG is Hallowed Halls of Gays To Go?
Re: (Score:2)
There are lots of infinities; surely the program at infinity-1 will be found before the one at infinity, but that does not in any way make it finite.
Even if you could narrow it down by 1/(the atoms in the Universe) you would turn out to be no closer to making it finite, or reachable in finite time.
Only enumerated numbers are reachable, non-enumerated fractions of infinity are still infinite.
Re: (Score:2)
Re:The Infinite Number of Monkeys (Score:4, Insightful)
But as a dev, I also do not have infinite brainpower.
An algorithm doesn't have to be perfect, it just has to be good enough to compete with human developers.
Re: (Score:2)
Shmaybe
Re: (Score:2)
It would be a waste of resources. What is does mean is software engineers will will do the engineering and the computer will write the algorithms. Same goes for animation, animation engineers will write an animation script and the computer will do the animating based upon an art palette, people and stuff, which would include music. So drunken drugged up minstrel are also in trouble as are the Muppet actors. Live performance will be the only revenue source left for them, engineers will gobble the rest up.
Re: (Score:2)
We've kinda had this for decades. Look at the way you could build a visual interface via drag and drop of a form container, buttons, input boxes, etc. in Visual Basic. Or how you can do it for Java with SceneBuilder. Or for Android with Android Studio. There is lots of boilerplate code to do things that can be computer generated/managed while the human just worries about what happens when a user actually clicks the button or changes an input field.
Re: (Score:2)
But for this since ages better approaches exist, such as http://nakedobjects.org/ [nakedobjects.org]
Re: (Score:2)
Thank the Gods for that one. The faster we can do away with "coders" the better off real programmers will be ...
Re: (Score:2)
With enough computing power AI will be able to write and test every algorithm possible.
Algorithms are but a fraction of app development.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2, Insightful)
The problem is the rate of job destruction (Score:4, Insightful)
The question is can we do a third industrial revolution without a century of war and misery this time?
Re: (Score:3)
more like 17M WW1, 60-80 WW2
Re: (Score:2)
I was just about to respond to something similar. However, the effect of WWII was that Stalin whacked about 20 million as did Mao. It is instructive that China reveres such a butcher, I guess covering up is easy for the CCP...as was the human rights mess they made of Tibet and are now perpetrating against the Uyghers.
Next up on the chopping block for the CCP: Taiwan. 23+ million free Chinese is not something the CCP can countenance lest the rest of their subjects start thinking they'd like to be free as wel
Re: The problem is the rate of job destruction (Score:3)
Re: (Score:2)
Sorry, my bad. Forgot where I was
Good direction for more investigation (Score:2)
"The problem is the rate of job destruction tends to outpace the rate of job creation." That's a good direction for more investigation. I think there are ways to minimize the problems.
"Forbes Technology Council is an invitation-only, fee-based organization..." [forbes.com] A "fee-based organization"? What does that mean?
Re: (Score:2)
A "fee-based organization"? What does that mean?
People who work for a living need not apply.
Re: (Score:3)
tends to outpace the rate of job creation.
America's population is ten times what it was at the beginning of the industrial revolution. Yet we have record low unemployment, and living standards have improved twenty-fold.
Folks just sorta gloss over the decades of unemployment that followed the industrial revolution
Can you name the decades you are talking about? The only decade when American employment fell was ... never. Even the 1930s saw an increase. And the decline in employment in the first half of the 1930s was not caused by automation.
Re: (Score:3)
Folks just sorta gloss over the decades of unemployment that followed the industrial revolution
Can you name the decades you are talking about?
That would be the 70 years between 1770 and 1840. And worker conditions remained poor in Britain for about four decades past that as is seen by shortened life expectancies and heights. There was a big turn around in living conditions from about 1880 to 1940, in cities anyway when advances in sanitiation, health care, food quality, transportation and housing standards were revolutionized.
The U.S. started to industrialize the better part of a century later than than Britain.
Re: (Score:3)
Additionally, until the U.S. had the advantage of being able to divert population into farming as the land available for cultivation expanded, reaching its limit in 1910. So the U.S. avoided the worst effects of the first stage of industrial revolution which cannibalized traditional employment faster than it produced other opportunities. Factory worker conditions remained very poor though, with wages below what farmers were able to make, and terrible conditions, into the 20th Century.
The conditions of the T
Re: (Score:2)
That would be the 70 years between 1770 and 1840. And worker conditions remained poor in Britain for about four decades past that as is seen by shortened life expectancies and heights.
Clearly the trick is to be born at the right time, so choose carefully!
Re: (Score:2)
Can you name the decades you are talking about? The only decade when American employment fell was ... never. Even the 1930s saw an increase. And the decline in employment in the first half of the 1930s was not caused by automation.
Exactly.
Ronald Reagan used to get huge applause when he would say, "America has more people working now than ever before!"
Well, duh. That's because America had more people now than ever before.
It's no different than saying "I've been alive longer today than ever before!"
Re: The problem is the rate of job destruction (Score:2)
Re: (Score:2)
Yes, that's what I was trying to point out.
My point was that Reagan, while technically telling the truth, was still being disingenuous.
The unemployment rate averaged 7.5% under Reagan. Declining steadily after December 1982, the rate was 5.4% the month Reagan left office.
So yeah, we had terrible unemployment numbers, but we still had "more people working than ever before."
I'm closer to death than ever before, but yippee, I'm also older than ever before. WINNING, right? lol
Re: The problem is the rate of job destruction (Score:2)
Re: (Score:2)
Going from 7.5 to 5.4 is *great*.
Sure, but it's still a terrible unemployment rate. Going from a 50% unemployment rate to a 30% unemployment rate is great, but again, it's still a terrible unemployment rate. I mean, yippee that it's dropping (that's great), but the direction is basically irrelevant if the end number is still horrendous.
Direction is as if not more important that raw figures.
No...going from a 50% unemployment rate to a 49% unemployment rate isn't great, even though the direction is good. In that case the number is more important than the direction.
As above, the direction is near
Re: The problem is the rate of job destruction (Score:2)
We're achieving that with tricks and lies (Score:2, Troll)
Go look up the history of the Luddites. The actual Luddites, not the term they're named after. Go look at the rust belt. Those jobs are never coming back since the factories are automated. Look up how the Steam controller was made, or how Gundam model kits
Re: (Score:2)
Re: (Score:2)
Folks just sorta gloss over the decades of unemployment that followed the industrial revolution until tech caught up and created more jobs
Probably because it didn't happen. Jobs were lost, and they were not replaced. Decades of unemployment followed, until the displaced workers had finally died.
Meanwhile, younger workers were learning the new jobs that had already been created; in fact it was the creation of these new jobs that was a necessary precondition for the decades of unemployment that you reference. You don't go from shops building things in small quantities, to nobody building anything, to decades later having factories. That isn't h
Re: (Score:2)
Re: (Score:2)
Yah, 'cause all automation is created equal, yes? Please tells us your analysis comparing previous examples of automation and the current one. You'll be wanting to reference scholarly articles on the subject. I think we can handle the facts, if you have any.
Actually, if you could really automate progrmaming (Score:5, Insightful)
The demand for app developers would go up, because the marginal productivity of people who could figure out what an app needs to do would go through the roof.
Re: Actually, if you could really automate progrma (Score:3)
I agree. If you could automate app development then the price of an individual app would drop and therefore the demand would go up. There are lots of people out there with ideas for an app. I get people approaching me all the time. Some of the ideas are actually good and could probably sell good in a niche market but these people canâ(TM)t afford the 20k entrance fee to find out. If instead, they could develop it themself or with limited help, that 20k app becomes a $2k app so you would see a lot
Re: (Score:2)
Funny how people have no idea how a modest team of people who know what they're doing can suck up $20K in a blink of an eye. You an buy a new car for that. Granted, a crappy new car, but still.
What people really need is for app development to be cheap enough that they can do things speculatively, or at least iteratively. People are too close to their own businesses to really see the problems in them or layout some kind of elaborate plan to improve their operations.
Re: Actually, if you could really automate progrm (Score:2)
Funny how people have no idea how a modest team of people who know what they're doing can suck up $20K in a blink of an eye. You an buy a new car for that. Granted, a crappy new car, but still.
Plenty of apps are written by a single developer. $20k can buy a single developer for a month or two which is enough to develop a simple app. If you truly had a no code system where a single developer can develop an app in a week instead then the demand would increase. Of course Iâ(TM)m skeptical of this because even in a no code environment you still need to design the layout, flow, graphics, etc... which to look good is going to require a designer even if you completely eliminate the programmer.
Re: (Score:2)
My old business partner used to get angry when he read about somebody who made huge bundle doing something like a ring tone. I told him he was in the wrong business. Just paying yourself to negotiate the contract and figure out what the customer needs is going to cost more than putting together a ring tone -- not to mention the support the customer expects from an app.
I don't doubt a single developer can churn out a $20K app, if it's not particularly critical and it looks a lot like another app he's alrea
Re: (Score:2)
The app price would not go down. Why would it? The costs to make one are the same! And customers can only bus so many apps per month anyway.
Re: (Score:3)
Indeed. This is not going to happen though. This must be the 6th or 7th time "automated coding" or "no-code coding" fails.
Re: (Score:2)
Re: (Score:2)
I will not worry even then. I mean, "end-user written applications" already exist, mostly in the form of excel sheets. When you do not need performance, reliability, security, maintainability, documentation, backup, etc. they can solve some problems. An actual expert in a different field can encode simple things in them. Yes, there may even be complex mathematics involved, but the coding side is exceptionally simple. This has not threatened competent software developers in any way. It does threaten enterpri
Re:Actually, if you could really automate progrmam (Score:4, Interesting)
The demand for app developers would go up, because the marginal productivity of people who could figure out what an app needs to do would go through the roof.
Indeed. This is Jevons Paradox [wikipedia.org]: When the efficiency of a resource (including labor) goes up, the demand often increases.
Think about it from the point of view of a greedy profit-seeking capitalist: If you could automate your production so each worker could produce twice as much, and generate twice the profit, would you fire half of them or hire more?
Re: (Score:2)
This is Jevons Paradox [wikipedia.org]: When the efficiency of a resource (including labor) goes up, the demand often increases.
The sentence, as written, could stand - if you did not link it to Jevons. Sure efficiency increases with increased usage - since the experience base increases, and there is increased incentive to invest in efficiency improvements. But Jevons was claiming that the efficiency increase caused increase usage rather than merely accompanying it.
To date there has been no convincing case made that Jevons paradox actually occurs.
Jevons himself did not show that it existed, but was merely speculating. His example was
Re: Actually, if you could really automate progrma (Score:2)
Easy answer. You fire half of them, outsource what you can, and extend the required hours of the remainder so you still get 2-3x the output for half the cost.
Re: Actually, if you could really automate progrma (Score:2)
Re: (Score:3)
Yeah, this is the actual answer. If the coders are already writing all the apps you can sell, and their productivity doubles, then you definitely fire half of them. If you can sell four times as many, then you're currently trying to quadruple your headcount, and instead you will try to double it.
Re: (Score:2)
Except demand will vary with prices and features. If productivity doubles and your costs are cut in half, you can reduce your prices or add extra features. Or your company could make new apps that were previously cost-prohibitive.
In the 1940s, IBM estimated that there was a worldwide demand for about five computers. But that changed when prices fell.
Re: Actually, if you could really automate progrm (Score:2)
I've been in this industry for 25 years now (Score:5, Informative)
I've been in this industry for 25 years now, and throughout all this time, pretty much continuously we were promised "low-code/no-code" development, and that it would make devs obsolete. Meanwhile in the real world we're experiencing record breaking demand for people who can design and implement complex systems, with code.
The world is becoming FAR more complex. (Score:4, Insightful)
The world is becoming FAR more complex, not less complex. For examples, watch Modern Marvels on NBC. [history.com] Some of the automation is amazing, but the results of the automation is that people want more.
Re: (Score:3)
Same here. When I was at university 30 years ago, the 5GL project failed, which had pretty much the same goal. And it was not the first attempt. The problem is that most people still have not the faintest clue how hard coding beyond extremely basic things actually is. Hence they thing the problem is just the language, when the language is only a minor part of the whole thing.
Re: (Score:3)
It goes back a lot further than that, to at least the 1950s with projects like AUTOCODE, back when computers were just a step above being programmed via plugboards.
What the people who make these predictions fail to mention is the rest of the prediction: The app developers will ride the monorail to work and use computers running off fusion power too cheap to meter.
Re: (Score:2)
If they're Japanese they might in fact ride the monorail to work.
Re: (Score:3)
I've been in this industry for 25 years now, and throughout all this time, pretty much continuously we were promised "low-code/no-code" development, and that it would make devs obsolete.
Exactly. I've been hearing this for decades, literally decades and we're not really any closer than we were before. Every time they try it they find out that there has to be a brain in the loop somewhere.
Re: (Score:2)
There are plenty of coding tools that make coding much much more efficient and faster than decades ago.
No idea why half of /. misses that and disagrees.
Re: (Score:2)
It's because everything that makes apps easier to create also allows skilled developers to do even more too.
Take the development of rich text editing controls. Suddenly anyone can have completed rich text editing in their app. But now everyone wants more than that, it's become the absolute minimum they expect and unless it's also got a spell checker they aren't going to buy that app. Just because a kid with Visual Basic can make a word processor in 5 minutes doesn't mean that is a valuable skill because the
Re: (Score:2)
Re: (Score:2)
Indeed. I've seen dozens of 'visual coding environments' come and go, and yet vim remains.
There was an article about the 'end of Moore's Law' I read recently that actually stated the opposite of this post. As die shrinks become increasing unaffordable and as Amdahl's law limits parallel efficiency, in a decade or so performant languages like C (and their developers) will be in greater demand than ever.
Re: (Score:2)
in a decade or so performant languages like C (and their developers) will be in greater demand than ever.
Unlikely. For everything that really matters, Java is as fast as C, and 100 times easier to use for multi threaded applications.
Lot's of luck with (Score:5, Insightful)
"everyone will become a developer". People aren't logical enough. People that can't code because they don't understand logic will be no better at describing their needs to a machine.
Re: (Score:3)
"everyone will become a developer". People aren't logical enough. People that can't code because they don't understand logic will be no better at describing their needs to a machine.
Yep, the real job of developers isn't coding, it is figuring out what the customer really wants and managing that with what the computer can really do.
Re: (Score:2)
I was channel surfing a few days ago and the issue was a pilot had something hit his plane. It wasn't clear what it was so naturally the subject of UFOs came up and, necessarily strange radar blips near his aircraft. They interviewed the fellow's sister as he didn't make it. Her ironclad reason it could not be pilot error: he was an experienced pilot.
With reasoning skills like that, I'm not too worried about citizens writing apps as long as they don't foist them on fellow citizens.
Coding is the disruptive technology ... (Score:2)
... to crafts.
It amazes me that a person can have an idea and go to Hobby Lobby and get a glue gun, some thread and paint, some wood or stuff and create a real-live version of their imagination.
I amaze myself, and am amazed by others, that I can code something from a thought to a practical (for me) application via code.
When coding is simplified to the point that those two approaches to art can meet, a whole ecosystem will flourish and be worth billions of dollars annually just in the US.
There's gold in them
Re: Coding is the disruptive technology ... (Score:2)
Re: (Score:2)
There's gold in them thar hills.
It's called pasteexchange, I think. Or stackbin, maybe? Something like that.
Just remember: Never borrow another man's Caribou board.
The view from 30,000 feet (Score:5, Insightful)
I think it is a fair bet that none of those "successful" people have ever performed any of the jobs they think will be automated away. As such their view of what those jobs require is at best second-hand and at worst totally disconnected from reality.
People have been talking about program generating software for what? nearly 40 years. So far none of them have made any sort of dent in the need for real programmers, solving real problems. While we get high(er) level languages, the problems, shortcomings and unforeseen limitations never seem to go away.
Though if we can get AIs to write some decent documentation, that would be a good start. That would require action from at least six of the 14 work areas that Forbes' team dreamed up, so I doubt we'll get it any time soon.
Forbes is "disconnected from reality". (Score:4, Interesting)
That's what happened at Boeing. Top management was, and still is, far from understanding the technical details.
Automate the CEO (Score:4, Funny)
We're more than half-way there [youtube.com].
Short answer: Sort of (Score:2)
Re: (Score:2)
Re: (Score:2)
A world where no thinking is necessary by humans? (Score:2)
"I think as workflow technology expands, any kind of oversight-related job will be delegated to the bots."
"In the next five years, everyone will become a developer thanks to low-code/no-code technology."
No thinking will be necessary by humans? We can all stay home and take naps?
The Forbes Technology Council [forbestechcouncil.com] apparently doesn't have anyone with technology experience to manage the council.
Reports (Score:2)
Yes, there are tools that will build reports for you. But it takes someone who knows how to build a report manually to use them. By the time you know how to get the data out of the tool, you know enough to write a report. Crystal reports (or whatever) costs a whole lot less than those tools.
No! (Score:3)
Betteridge!
CASE (Score:2)
Didn't work then, doesn't work now.
Depends on what you mean by "app" (Score:3)
If what you call an app is essentially a static document in app form, like many apps are, then I don't know why there are people who code that even now. Code simply isn't the best way to design a nice layout, and that's what you need for such an app. The rest is notifications, but again, you don't really need code for that, just connect two dots.
But some apps are real programs. Games, even those that use an engine, need some logic. Even messenger apps need some code to process user input and deal with a lot of dynamic content. It can be web based but there is still code. Some apps do image processing, automation, navigation, etc... Decades of experience of trying told us that low-code/no-code has limits. Good for doing some very specific things, but if you want a bit more freedom, code is just too powerful, at least for now, but it shows no sign of stopping.
Re: (Score:2)
Hopeful next step: (Score:3)
Automated scrum meetings ...
Re: (Score:2)
Re: (Score:2)
macOS, iOS, all those are based on NeXTSTEP, and many of Apple's bundled applications started there too.
Their fans claim that development is simpler there. Certainly since the interface is simplified and most of the applications do less, there almost must be some sort of truth to it. If the benefits scale to full-featured software isn't as clear.
You can never automate the PHB (Score:2)
I want a report that has xyz. Here you go. No I meant XYz. Oh, here that is. This report is incomprehensible, what does X and Y mean?
No way (Score:3)
---
> In the next five years, everyone will become a developer thanks to low-code/no-code technology. It allows users to build apps and processes in a visual integrated development environment with drag and drop features. Hand-coding isn't likely to become obsolete in five years, but we are moving towards a far future where little to no coding is involved in development.
---
I don't believe this at all. Programming has always been and always will be hard. I've seen very little transition away from c# and java stuff. The US markets are presently in a virtual recession. The big companies still dictate much of the terms of enduser software. What they're envisioning with automation of all code is probably more than a decade away. Even the utility "Chef" takes technical know-how to run efficiently.
Low-code tool maker says we all need low-code tool (Score:3)
Maker of a thing says we will all be using her thing in five years? I'm shocked, shocked I tell you...
sure (Score:2)
Oh yeah, any day now.
But how are we gonna describe the business logic to the robots? Maybe we can come up with some kind of symbolic languages to represent it ...
The Stupid, it burns, it burns! (Score:2)
I remember that. (Score:2)
Has anyone used these? (Score:2)
Has anyone used these low/no-code platforms? Actually I'd be quite interested in using them. I'm an app developer and although I've done plenty of backend stuff in the past, I've grown a bit out of it. If I could simply click together a bunch of tables that came with a REST interface, I'd be quite happy.
Everybody is a programmer - not (Score:2)
Low-code/No-code, again...
If we go back a couple of decades, this is what Excel macros enabled. We see exactly how well that works: Sure, everybody can write a macro, or put some really abstruse calculations into a spreadsheet. My office has a couple of people in admin who love Excel sheets, and think they are geniuses. I spend more time trying to force information into their overly-clever sheets than I used to spend on a simple form. Time savings is negative.
Another example is SAP: They have a configuratio
Is this the year of the Linux desktop? (Score:2)
Re:Dreamweaver. (Score:4, Funny)
Re: (Score:2)
I've used a few generations of those types of tools, and they always require way more clicking and dragging than just typing in the code, if you know what the app is supposed to do... and if you don't, you click and you click and you click, and in the end it does... something? Or at least, more and more nothing. That eventually equals something, right? If you click long enough?
There will always be people who don't know how to program, who can take a list of requirements, and operate the "code-free" app buil