Say you're given the full-time labor of the world's most capable software engineer and computer scientist for a short period of time, say 1-3 years. The only flaw: he has precisely zero creative capacity outside of the task he's given.
What do you have him develop or invent for you that would provide the greatest payoff?
Machine learning algorithms for finance. Or bioinformatics algorithms for faster genome sequencing.
>>1667061
Specifically which parts of finance? A lot of the big players are currently investing a lot of money into machine learning for finance.
Design an program that collects all world economic data and produces accurate data on stock price forecasts and autonomously leverages it to short General Motors.
>>1667067
Is this data available anywhere?
>>1667066
Yeah it's the current hotness. So, in this case, I would say have the guy make something that can heuristically guess other's traders positioning or take positions based on market movement - basically an algorithm that feeds off other's actions rather than underlaying securities,
>>1667053
I need a program that allows me to in real time alter the genes in trillions of human cells. I need the software to control the process with latency being the biggest problem.
>>1667072
That's interesting! Do you already have ideas on potential sources to use as a live stream of data for such a system?
>>1667077
This is kind of what I was getting at. Gene data and real time are not going to happen with current tech. Data sizes are too large to efficiently solve without a breakthrough algorithm, or cutting the problem space down and working with probabilities. My pet theory is quantum computations could solve this problem w/o latency if you structured it right. Sadly even quantum computers right now are far from useful. Is yours actually a real problem or just hypothetical to this thread?
>>1667053
A smartphone fart app