Dwave Discusses Building A Big Idea Company and Speculates on Super Intelligence

Dwave Systems is trying to build a quantum computer company.
There is controversy about whether Dwave has sufficient quantumness in their qubits. Most of Dwave System insights into making a company based on groundbreaking technology still applies regardless of whether Dwave technology succeeds. These insights would apply to molecular nanotechnology and artificial general intelligence.

Phase I (5 years)
Formed in 1999 to commercialize quantum computers
• Entered into research collaborations with many institutions
• Learned a lot on a budget
• Filed a lot of patents

Dwave’s first business model was inadequate

First business model was helping existing academic QC (Quantum Computer) efforts in exchange for IP (Intellectual Property) presupposes those efforts will build QC in time frame relevant to investors

They (Academics) Won’t …..
* not motivated to
* not capable
* never have
* wrong culture

-DWAVE
must create infrastructure internally – address current limitations to progress with superconducting quantum computing adopt practices of successful technology development startups identify and address issues currently limiting progress in field
– primarily technical/fundamental or organizational

Phase II (5 years)
• Chose a particular technical approach
• Pulled (almost) everything in under one roof
• Serious $$$ (order ~ $100M)
• Infrastructure building (fab), de-risking basic science

Phase III
• Selling current systems
• Operations plan built around sales revenue
• Need for sales & sales leader
• Cultural shift from inward focused PhD science & engineering to customer facing R&D
• Technology works: a double edged sword!

Dwave Speculates on SuperIntelligence

quantum algorithms exist for solving hard learning problems
… machines running these algorithms might be superior to any possible evolved brain
• Pattern matching
• Inference
• Deduction
• Scheduling
• Optimization

Why don’t we have super-intelligent machines already?
• No-one has figured out how to (“scalably”) generalize learning
• “Input!!!!”