Interesting! So the cloud wil be the invisible infrastructure that’ll support unlimited memory? How about cheap universal 3D imaging devices that’ll replace 2D? Do you foresee progress in brain-computer interface? Something’s gotta replace the kbd, the oldest component of all personal devices, so voice input?
Christopher Smithsays:
Don’t forget how fundamentally CPU design is expected to change. We’re going to have memory embedded with each core, and the cores will likely talk to each other through a network stack.
So the cloud wil be the invisible infrastructure that’ll support unlimited memory?
The cloud already provides essentially infinite storage. However, in 2015, programmers have to worry about preserving memory, especially on mobile device. I think that this is a temporary limitation.
Of course, nothing is ever really infinite… by “infinite”, I mean “something you do not worry about”. (E.g., you have access to an infinite amount of popcorn.)
How about cheap universal 3D imaging devices that’ll replace 2D?
I’d be more excited about direct retinal or brain projections.
The eye is not nearly a perfect interface, if we could bypass it, it would be far better. Think of all the people who need glasses to read? Why would you need glasses to interact with a computer? Why should bad optics be a problem for computer interaction?
But I sure hope that surgeons will have access to full 3D projections of your body prior to critical interventions by 2025… but maybe that’s too hopeful… since doctors still have stethoscopes from the IXXth century today.
We are being promised full virtual reality in gaming for the 2020s… but it seems we are still a full generation of two away… I guess that in 2025, we might get the PlayStation 6… who knows what that might do?
Do you foresee progress in brain-computer interface?
Handicapped people can control a mouse using their brain today… the problem is that it is invasive and not quite as accurate as a real keyboard.
We already use brain-computer interface to try to help people with cognitive diseases. Results are not quite conclusive yet…
I don’t know…
Something’s gotta replace the kbd, the oldest component of all personal devices, so voice input?
I find myself increasingly tempted to use Siri or Google voice search. It is not yet good enough for regular use, but if we can bring down the error rate by a factor of 4…
Would you want to do everything with voice though? Probably not…
I agree with you that we have to replace the keyboard at some point… Brain-computer interface would be *the* thing, but I don’t see it for 2025…
So maybe we will still be stuck with keyboards. I hope not.
Ksays:
”So maybe we will still be stuck with keyboards. I hope not.”
How about devices like leap-motion and neonode airbar?
Don’t forget how fundamentally CPU design is expected to change. We’re going to have memory embedded with each core, and the cores will likely talk to each other through a network stack.
This is now around $3, has built-in wifi and can run a node.js-like server written in Lua that can talk http or mqtt.
At this rate, I expect to see at least 16 cores, 1 gig of RAM and built-in wifi / bluetooth for under a dollar by 2025. Such these things will automatically mesh into a supercomputing cluster whenever you bring them in range of each other and Elixir will be the language of choice for learning to program them.
Interesting! So the cloud wil be the invisible infrastructure that’ll support unlimited memory? How about cheap universal 3D imaging devices that’ll replace 2D? Do you foresee progress in brain-computer interface? Something’s gotta replace the kbd, the oldest component of all personal devices, so voice input?
Don’t forget how fundamentally CPU design is expected to change. We’re going to have memory embedded with each core, and the cores will likely talk to each other through a network stack.
@Stan
Congratulations on the job at Dalhousie.
So the cloud wil be the invisible infrastructure that’ll support unlimited memory?
The cloud already provides essentially infinite storage. However, in 2015, programmers have to worry about preserving memory, especially on mobile device. I think that this is a temporary limitation.
Of course, nothing is ever really infinite… by “infinite”, I mean “something you do not worry about”. (E.g., you have access to an infinite amount of popcorn.)
How about cheap universal 3D imaging devices that’ll replace 2D?
I’d be more excited about direct retinal or brain projections.
The eye is not nearly a perfect interface, if we could bypass it, it would be far better. Think of all the people who need glasses to read? Why would you need glasses to interact with a computer? Why should bad optics be a problem for computer interaction?
But I sure hope that surgeons will have access to full 3D projections of your body prior to critical interventions by 2025… but maybe that’s too hopeful… since doctors still have stethoscopes from the IXXth century today.
We are being promised full virtual reality in gaming for the 2020s… but it seems we are still a full generation of two away… I guess that in 2025, we might get the PlayStation 6… who knows what that might do?
Do you foresee progress in brain-computer interface?
Handicapped people can control a mouse using their brain today… the problem is that it is invasive and not quite as accurate as a real keyboard.
We already use brain-computer interface to try to help people with cognitive diseases. Results are not quite conclusive yet…
I don’t know…
Something’s gotta replace the kbd, the oldest component of all personal devices, so voice input?
I find myself increasingly tempted to use Siri or Google voice search. It is not yet good enough for regular use, but if we can bring down the error rate by a factor of 4…
Would you want to do everything with voice though? Probably not…
I agree with you that we have to replace the keyboard at some point… Brain-computer interface would be *the* thing, but I don’t see it for 2025…
So maybe we will still be stuck with keyboards. I hope not.
”So maybe we will still be stuck with keyboards. I hope not.”
How about devices like leap-motion and neonode airbar?
@Christopher
Don’t forget how fundamentally CPU design is expected to change. We’re going to have memory embedded with each core, and the cores will likely talk to each other through a network stack.
Right. I agree that it sounds plausible.
In 2025, I personally intend to be programming in Clojure (assuming that everyone else is on JVM / Javascript VM)
I hope there’ll be a fairly syntactically close native Clojure compiler so I can dispense with the C, too.
The most exciting thing I’ve seen recently is the http://makezine.com/2015/04/01/esp8266-5-microcontroller-wi-fi-now-arduino-compatible/
This is now around $3, has built-in wifi and can run a node.js-like server written in Lua that can talk http or mqtt.
At this rate, I expect to see at least 16 cores, 1 gig of RAM and built-in wifi / bluetooth for under a dollar by 2025. Such these things will automatically mesh into a supercomputing cluster whenever you bring them in range of each other and Elixir will be the language of choice for learning to program them.
@Phil
I expect to see at least 16 cores, 1 gig of RAM and built-in wifi / bluetooth for under a dollar by 2025
Yes.
If your prediction is correct then GPUs either will be replaced with CPUs or gain even better performance 🙂
We will probably not be having a separate CPU/GPU. We will be having IPU’s instead. Integrated processing units. I just coined the term 🙂
@Jafar
I think that the distinction between GPU and CPU is already blurring with technologies like Knight Landing.