What will be the top 3 trends in 2016, that shape how we build our startups?

[aszig]

Hey there,

During the holidays I was thinking about how 2016 will be different, better than 2015. There are a lot of articles about emerging markets, technologies, but those are the focus of entrepreneurs. But I'm interested in what will change in how we build startups...

Here are my top 3 guesses.

1. Augmented productivity: I'm a fan of the gig economy ever since I ordered the first $5 logo design from fiverr. And I'm just starting to explore MTurk. This year I'll put a lot of effort into learning how to combine these with workflow automation services like Zapier, and new AI stuff... I think such methods could multiple personal productivity. Inspired by the guy who automated 90% of his work. http://www.businessinsider.com/programmer-automates-his-job-2015-11

2. Improved startup-building platforms: It would be such a blessing to multiply the ratio of successful startups, and lower the pain of failing. I mean, 9 out of 10 startups fail even in developed startup hubs. That's one metric that we need to fix. Traditional incubators and accelerators are nice, but not too effective in mitigating risks. I loved how YC started Research (http://blog.ycombinator.com/yc-research), and even more, that there are now so much more startup studios (http://buildtogether.co/startup-studio-list), and finally there's a booklet (http://laicos.com/anatomy-of-startup-studios-is-out/) that explains in detail how these venture builders work.

3.Crowd*: Besides perk based crowdfunding, there are more and more platforms where you can offer equity in your startup for support. That just adds to the option of the founders to get the money they need... who know, maybe even faster than angel money and with better than vanilla term sheet conditions.

What do you think, what are your picks?

@aszig

[tmaly]

Here is my guess

1. More AI bots and machine learning applied on top of apps etc.

2. Less money for startups while they are private, and more pressure to go public to handle the liquidity issue and valuations we saw with Fidelity

3. More platforms and infrastructure automation to help with the algorithms and cloud side of the equation. Google recently released a deep learning api that runs in the cloud. I think we will see more of this cloud driven AI/algo type services.

[sebg]

Chat bots come to life through Slack and other platforms

[k__]

Also they will be sponsored by Uber, who adds deep learning to all the bots, so they figure out where and when you need a cap. This will allow them to apply special on-the-spot surge pricing moments before you call them.

[cbzehner]

Skynet.

[aszig]

The long term future is not human vs machine armageddon, but human-machine integration. We are already cyborgs with our addiction and dependence on smartphones (not to mention ppl with pacemakers, prosthetic limbs...)

Actually, it would be kinda cool to plug into the Machine. So you don't have to type and click to give instructions. Just think about it and the Machine Spirit will make it happen.

But I think this one is not for 2016 :)

[DrScump]

Samsung refrigerators become self-aware, demand glacial water

[coderKen]

AI will be the new UI. before you downvote ask me for an explanation.

[aszig]

Please explain? :) Do you mean some Jarvis (or Watson?)-like stuff that can recognize natural language, and figure out how to solve?

[coderKen]

Definitely, you see the use of buttons and textboxes and checkboxes will slowly become obsolete, especially as advancement in natural language will continue to rise. So I just got a new Android Phone and since I got it, I have never for once set an alarm or reminder manually, I just do the "Ok Google" thingy and talk to it and boom, it's all done. This made me realize that the user-interface as we know it today will slowly become obsolete and even more so when AI is embedded in the OS and an API is exposed so that third-party apps will be able to tap into it.

Let's Imagine a future where the Android OS has become so advanced say v9.3 and has natural language API embedded that apps like say Uber can tap into, if you wanted to request for a ride, you wouldn't need to tap silly buttons and drag a pin across a map, you would simply say "Hello Uber I am going Trafalgar square, kindly request an Uber X to come pick me up"

PS: I build android apps

[aszig]

"I just do the "Ok Google" thingy and talk to it and boom, it's all done. "

I wish such apps (incl Siri) would be able to handle accent. Sooner than later, I hope.

[EvanPlaice]

1. Bots become more prevalent on the business end

Developers and designers work best when they're not interrupted. Asynchronous communication enables workers to ride the flow when they're focused and address communication when they take a break. A bot that requests a daily status, collects results throughout the day, and summarizes them would save the time wasted in unnecessary meetings without sacrificing communication.

Managers tend to piss off their workers because they perceive their time as more valuable and expect their workers to drop everything for a status update. Bots have no ego, can collect all the necessary data, without being intrusive.

That's just one example. Why are we still paying people to manually aggregate data? Bots can and should be used for menial/repetitive work.

I think startups will start to popup to 'spread the bot love' to our less-technical business counterparts.

2. ChatOps and greater cross-platform integration

Microservices are on the rise, everybody is talking in REST, and ChatOps is becoming the backbone of successful organizations. Slack is a good example of this. What it can't integrate with by default it can be extended with using bots or glue services like Zapier.

It's time we ditch the push/pull models of communication. It doesn't scale and it's too easy for bad actors to game the system. Work, progress, and status should be tracked passively and in the open. No more lazy people hiding from their responsibilities, no more bad actors advancing by kissing ass and taking credit for other's work.

Many sites/services have been exposing public APIs with a 'build it and they will come' mentality. It's about damn time we tap those resources.

3. Unikernels will start to be used in production

For stateless webservices, it simply doesn't make sense to use a full-featured general purpose OS. A unikernel drastically reduces the surface area for security vulnerabilities, requires zero maintenance, requires very little memory/space overhead, and is small enough to fit in version control.

Docker containers will live on as a scaleable approach for 'pets', unikernels will start to emerge as a replacement for 'cattle'.

To take this a step further. If we're running a service that doesn't require mutable state, why the hell are we running it on a mutable operating system?

DevOps is expensive and difficult to setup. Organizations that figure out how to embrace this new approach will be able to move fast and not break things..

-----

Bonus: Anti-trends

1. TDD will be on the decline

Devs start to question whether TDD is just another form of 'waterfall' model. If we collectively agreed that making a ton of assumptions up front before starting development is a bad idea, why has that practice been pushed into the testing domain.

The important question to ask isn't, "how much test coverage do we have?" It should be, "how many tests are we throwing away?"

On a short timescale, startups will start looking to 'trim the fat'.

2. We reach 'peak mobile app'

Apple has been coasting on the success of mobile since the passing of Steve Jobs. They're betting the future on mobile at the expense of everything else.

Safari has become the new IE and the updates to OSX have been lackluster. OSX is the preferred platform for a large percentage of Developers/Designers and nothing pisses them off more than watching their favorite tools/ecosystem fall by the wayside. Resentment toward Apple will grow.

To make matters worse, both React and Angular2 are aggressively pushing to be the platform of both the web and mobile. UI responsive issues have been resolved with the VDOM, and they've effectively built architectures that separate the application from the UI rendering layer.

Mobile has never been scalable from a business sense. There's only a limited amount of screen real estate for apps. Mobile development strongly favors the few at the expense of the many. Plus, proprietary platform lock-in sucks. The mobile app ecosystem will contract as well as any -- but the most successful -- startups that depend on it.

3. 'Nobody ever got fired for purchasing Oracle' will become a past time

Amazon AWS is aggressively pushing to eat Oracle's lunch. With the recent launch of services that achieve feature parity with Oracle's db as well as tools that assist the migration, Oracle will be on the decline. Not to mention, the business world finally overwhelmingly accepted the cloud as the future in 2015. Amazon AWS is the sleeping giant that just woke up holding a can of whoop ass in one hand and a can opener in the other.

Startups that specialize in data migration to the cloud will be flush with cash.