In my opinion, the notion of an "algorithm for generating new ideas for startups" is absolutely appalling. I am far from an expert in the field of startups but I have enough knowledge (at the very least from a consumer's point of view) on the matter to know that great products are crafted by imagination and passion. Having a process decide what a "good startup idea" would be completely takes away from the base imaginative thought. Furthermore, how are you to become passionate on an issue that was brought to light by a process rather than intuition?
Is it really that bad though? I see it as a form of brainstorming and/or a way to overcoming blind spots/biases in your thinking. It's not an 'automated process' to generate ideas - just one method in a whole 'toolkit' of methods to stimulate one's creativity.
I can definitely see the benefits of using this tool as a brainstorming function, however why try and force ideas to create a startup. The function of a startup is to make a company in order to respond to a need. I believe that need is only going to be fully satisfied when the company/person/startup addressing it has some underlying attachment to the core issue. Grabbing a hold of an idea that is a result of some exterior process, to me just seems wrong and lacks a genuine intention.
If you believe that the human brain is entirely physical, then there's no reason why an algorithm couldn't output equivalent or similar results as an imaginative passionate human.
The claim that the human brain is entirely physical in its behavior? That's an extremely common (albeit contentious) viewpoint held by AI optimists, as well as many general humanists. (In fact, to argue against it is by definition to argue for the existence of the supernatural.) I have always assumed that the vast majority of computer scientists would believe that the brain (and indeed, the Universe) is computable (in theory, of course—we obviously would have trouble creating a computer big enough to actually do it). The popular CS books "Gödel, Escher, Bach" and "The Lifebox, the Seashell, and the Soul" both discuss the idea a lot, and the latter claims it as its central hypothesis.
> In fact, to argue against it is by definition to argue for the existence of the supernatural.
There was a relatively recent HN submission regarding a proposal that our memories are stored in molecules that outlive synaptic connections (more or less). One commenter pointed out that the principle author advocated a theory which says that at some deep level our brains interact with quantum mechanics and that this is the source of our consciousness.
Therefore, it seems that there are 3 possibilities depending on your beliefs - physical, quantum, supernatural.
I tried to find the article, the author, and the comment, but my Google Fu is weak when it comes to searching HN.
Quantum is physical. Just because some popular writers don't understand the math behind it doesn't turn it into a mystical "get out of causality and reason" card.
Whether the human brain is entirely physical is for a philosophical debate which I am not interested in.
I was merely interested in "there's no reason why an algorithm couldn't output equivalent or similar results".
The truth is, we don't know yet.
I personally believe in what you claimed.
I was just asking for any papers that supported your claim, because I would be really interested in those.
Sure, great products are crafted by imagination and passion, but the decision to create the product at all is usually based on analysis. Making some of that analysis automated isn't really such a bad thing, is it?
I think this is a proactive approach rather a reactive approach, when products/service come into existence after mass uses new technology and find issues with it and market respond to that need.
Another way to work with ideas is to work on the technology stacks behind them. Want to implement a new kind of service but the underlying technology stack doesn't work well for the idea? Work on that first, possibly in such a way that it can be sold or leveraged.
For example, Ruby on Rails came from Basecamp, SproutCore came from Mailroom, and Redis came from LLOOGG.
