Volume 34 Number 2
Six More Weeks of What?
By Michael Desmond | February 2019
A year ago in this space, I drew inspiration from the odd cultural phenomenon that is Groundhog Day to offer predictions for the year ahead. With magical, weather-prognosticating rodents once again seizing control of our climate, I thought it a good time to return with a few fresh predictions. And once again, I reached out to the team of contributing editors at MSDN Magazine for their vision of the future.
Last year, machine learning (ML) and artificial intelligence (AI) were top of mind. Both Test Run columnist James McCaffrey and Artificially Intelligent columnist Frank La Vigne predicted that enterprise developers’ roles would expand to include coding for AI and ML, much the way they did for mobile app development and DevOps. La Vigne says 2019 will be the year that AI hits fast-forward. “Last year, I saw a few technologies go from published paper to deployed service within six months. That kind of unprecedented pace is both exciting and terrifying.”
Keeping pace with fast-changing technologies is old hat to developers, but they’ll have help from automated ML solutions that streamline mundane tasks, La Vigne says. Longer term, he says evolving technologies may even “automate away much of the work that keeps data scientists gainfully employed.”
McCaffrey in his role at Microsoft Research has a front-row seat to ML developments. He expects rapid uptake of the PyTorch neural network library through 2019. And he predicts that enterprises will apply neural network-based prediction code to increasingly complex challenges, such as to sentiment analysis and to hybrid systems like deep neural networks combined with reinforcement learning systems.
Beyond AI, McCaffrey says, “I'm cautiously hopeful that in 2019, big progress will be made in the field of homomorphic encryption—a form of encryption that allows computation directly on encrypted data.” While such systems work in theory today, they’re still too slow for practical use.
Thomas Hansen is a regular contributor to the magazine, and a guy completely unafraid to articulate his vision of programming. He posits that new tools will emerge to create a fusion between the server and client, and offers as evidence the GraphQL query language for APIs and his own, home-brewed Lizzie scripting language.
Along those lines, Ted Neward (of The Working Programmer fame) says innovation around WebAssembly (WASM) will bridge gaps between disparate server and client code. Accessing code written in another language for a single-page application means implementing an HTTP-based API channel for the browser to call that other-language code on a server. It gets involved. Now, says Neward, “I can cross-compile my code to WASM, embed it into my webapp, and run everything out of the browser.” Neward says 2019 will see the rise of language UI libraries and frameworks using WASM to let developers write browser-based UIs in a higher-level language, while getting away from traditional DOM-based browser controls.
The Darkest Timeline
Last year’s predictions included concerns around attacks like Sceptre and Meltdown. Now concerns shift to impacts from AI. La Vigne says autonomous vehicles will eventually threaten to push millions of truck drivers out of work. In the nearer term, he says keep an eye out for DeepFakes, which take Fake News to the next level by using neural networks to seamlessly map the image of one person onto another. La Vigne thinks 2019 will produce “the first political scandal brought on by a DeepFake.” That’ll be fun.
The most troubling prediction may come from David Platt, who thinks his Don’t Get Me Started columns could be used to train an AI-powered rant generator. His dark vision for 2019? “A rogue TV commercial triggers home speaker voice AIs to start shouting at each other,” Platt says. “Riot spreads until the city cuts electric power.”
Now that’s the darkest timeline.
Michael Desmond is the Editor-in-Chief of MSDN Magazine