Data science. We are generating ridiculous amount of data right now, most of them are kept somewhere in a data lake. Companies don't have a clue on how to process or analyze them. I think in 5-10 years, it'll be a tipping point where the companies finally put a lot more money into that space.
My bet is on large-scale databases and data engineering. AI is hyped, but the sad truth is that a lot of the recent advances come from bigger models and making use of more data. Historically the same thing has happened a few times [0]. It's no different this time. Recent advances like big transformers, self-supervised learning, etc, are more examples of that. Even the initial convnet breakthrough in 2012 was possible only because we figured out how to process a large dataset efficiently on a GPU.
Companies will realize that they don't yet need Data Scientists or AI researchers. They need engineers that are really good at building data pipelines, event stores, feature stores, and databases specific to various query patterns. Or perhaps one to rule them all.
[0] http://www.incompleteideas.net/IncIdeas/BitterLesson.html
AR/XR will be one, check out the MixedRealityToolkit and the experience in the HoloLens 2
- Cheap small scale electronics: imagine raspberry pi-like devices with a sticker form factor, and very cheap, energy efficient enough to be powered by the sun. Can't think of a useful use-case, but kinda cool to shrink things like that
- Differential privacy
- Quantum computing
- Programs that can replace programmers
- ML for biotech
- Automated fabrication
From a more social perspective, in order of least to most abstract:
- Software that helps improve public infrastructure, logistics somehow - eg identifying ways to minimize small amounts of energy waste on a wide scale
- Tools that help local municipalities and citizens cooperate together
- Tools that give the disenfranchised a greater voice
Difficult to say, cause the perceived value of blockchain and ai is way higher than their actual value (see cycle of hype).
Maybe Quantum Computing will have its glory days in 10 years...
Biology+computing
Brain machine interface
Lasers into eyes (for eyesight improvement, and VR/AR)
Healthspan and lifespan extension
Dental care (for all the old-young people)
Space mining, manufacturing and construction
DS/ML/DL/NLP/NN/RL/AI. We are Just scratching the surface only right now. Application to Numeric, text, images, voice, motion inputs/outputs are still the niche and siloed.
Think of applying existing Techniques and developing new techniques to combination of Variety of inputs and generating combinations of variety of outputs.
I'm hoping for explainable ai. Here's this cool technology that can generate pictures of cats! Cool! How does it work? Uhm.... Someone needs to take all these millions and billions of parameters from neural nets and turn them into something humans can understand.
Also, electric velomobiles!
That's the big question, which can be generally only answered in hindsight. This is because the hypes are largely irrational and who knows which fairly small/inconsequential idea will gather a ton of mindshare and which one will be ignored.
It is always better to be on the train.
Select the train according to your likings. -- VR/AR -- Lab-grown food -- Digital Twins -- Biotech -- Personal Moats -- Longevity -- etc...
Citizen science, that is making people feel smart while collecting / generating huge amount of data and working together toward a bigger social goal.
probably within the set of whatever research has been funded by DARPA or IARPA in the last five years
Hardware & Electronics
Medicine.
Whatever Musk is doing
It's still going to be AI/ML. The AI summer started with convnets, then generative models like GAN, and more recently transformer-based NLP. Each of these represents a step change in capability, fuelling its own wave of hype.
It's impossible to forecast these types of breakthroughs, but they'll keep happening. My bet is that in 10 years we'll have extremely life-like agents that come close to passing the turing test.