My Reinforcement Learning metaphor for equity finance driven startups is that they are the exploration process for capital seeking perfect ROI. After they find a new high margin process they exploit it to maximum financial return and (they hope) create a new paradigm that placed them as a monopolist or at worst oligopolist.
I’d argue this is precisely how major startup financiers view this - thiel explicitly says this.
So given that, and given that technology often is a key differentiator for structural change through new processes, it makes sense that evaluations of regulatory capture processes (including patents) would accrue primarily to new entrants.
Worth noting though that if you further iterate this capital/tech game, other larger monopolies/oligopolies will accrete these companies and technology too until some social tipping point is reached and the whole thing collapses.
I discuss this process as it’s root form in the third proposition of my myth of scarcity paper [1]
I like the RL metaphor. Does this imply a corollary that incremental success by exploration is more probable in new spaces but gradually decreases as those spaces get explored?
If you think of a three dimensional manifold as the unexplored state space of a new market, then the exploring:exploiting process maps areas of the manifold into “known” and “unknown”
It takes effort to explore the manifold in real life and the effort is measured in $/hour. Meaning whomever shows up to a new unexplored manifold with the highest potential work energy will be able to explore/search the space faster unless someone else has a more efficient search/explore or exploit mechanism.
I think this would be where “tragedy of the commons” would be relevant
Can you explain the tragedy of the commons connection? Are you implying it erodes when there is a lack of coordination and eventually provides sub-optimal outcomes?
It important conceptually to align "commons" with "new markets" because in practice they look the same with respect to consumption. Fishing is my favorite theoretical example here.
Assume you are a fisherman looking for fish in the ocean
To take an exploration "step" requires work: Acquiring a boat, acquiring fishing tools, having the energy to fish, having time to explore the ocean etc...
Once you find a cache of fish, then you switch to exploit by harvesting the fish from the cache
This assumes that the fish are not currently the property of someone else, and therefore it is "commons" or a "new market" until claimed by some entity with enough power to exclude others from consumption.
The problem is, information in new markets is sparse. By definition you cannot know prior to exploring it, what the size and scope of this fish cache is. However consumption starts PRIOR to the evaluation of the cache and what an appropriate sustainable exploitation rate is because doing so would assume something else: That there is an evaluation of a global optimum for that fish cache as a common resource.
Almost no organizations have the incentive to do the above, and history shows that most consumption orgs have narrow near term goals of maximum exploitation.
So the groups with the largest amount of energy to expend exploring and then exploiting (aka how much money do you have), will by default monopolize the resource.
Thank you for taking the time to explain. I think my initial confusion comes from conflating the "commons" and "new market without ownership." I don't think they are necessarily the same in terms of the "tragedy of the commmons."
I don't see why the "commons" has to have anything "new". Take the example of air pollution. The air is a common resource that people have access to without cost, but it's nothing new. So the predicate that "information in new markets is sparse" doesn't apply.
>"Almost no organizations have the incentive to do the above"
I would argue that collectives, like the government, have this incentive. Hence the need for regulation.
There is nothing on the market structure stopping consolidating companies to commercialize new tech by themselves. In fact, if you take a naive view of the market, the consolidated companies are the ones with all the structure needed for it, while startups have all kinds of obstacles to get a foot on the door.
This is a well known fact, found again and again by several studies. But it is not obvious in any way.
This is similar to the point being made in "The Innovators Dilemma" [1].
I experience this currently at the wealth management bank I work. They were reluctant for a long time to invest into a mobile app for trading because it may cannibalise their current portfolio. They built it last year, but pricing was close to our bespoke services. The developers built a great product but when pricing is double whats on the market, then you are going nowhere.
The question "Will it cannibalise our current profits?" is asked to stop every idea. Rightfully so maybe: We make 750k revenue per employee. Money is bursting through the seams. Shame it all goes to shareholders, not employees.
> The question "Will it cannibalise our current profits?" is asked to stop every idea.
I wonder if this is the real objection, or really just a proxy one? As in, yes, the new thing will compete directly with the old thing, but it will easily bring in way more than it "cannibalizes". So the company will likely come ahead. However, the old thing and new thing are likely developed by distinct teams, in different business units of the company. So I imagine this question is really a proxy for the old thing team asking, "will it cannibalize our currenet jobs?". If the answer is yes, it's not surprising they'll resist.
I assumed it's also an easy trap to fall in to if you start to write down pros/cons in a list when you discuss / strategize where to invest your next effort to increase profit. Effort to optimize current thing, have a lot of pros and few cons, so very low risk. Something that could cannibalise current profit might have few bullet points at the pros list vs the cons, even if that one bullet point might be substantially bigger in the effect.
Sony had the technology(mp3 players, phone, etc) and competence to be the first one to invent the smartphone. But I can't imagine in a risk averse culture, that a boardroom would listen to someone that would come in a suggest something that cannibalise multiple divisions in the company, in the promise that it would generate more profit, vs each divisions proposal of how to optimize their already existing profits.
Palm and blackberry got crushed by apple, who weren't the inventors of the smartphone but rather the ones who came out with a product that could be successfully marketed for over a decade. The smartphone is very different from a lot of the software projects people discuss here in that there are a TON of moving parts to release something as functional as the iPhone.
It is an uncomfortable risk for many, to back a likely long term winner over a sure current cash cow. Even if their own estimates say they should go for it. See it all the time.
And to make it even more short sighted, the idea won't go away, it will just take a little longer and then pop up in competing colours.
If you care about short-term profits it makes sense to ask that question, if you care about long-term profits it doesn't.
I think it's going to be very rare that over the life of a business every opportunity will be orthogonal to your existing business lines and not cannibalize them in some way.
Companies seem structurally incapable of cannibalizing existing business lines even if it barely impacts revenue because they cannot afford to report reduced growth for even a single quarter.
The difference is that employees at startups get a larger share of the long-term profit. At a megacorp, you mainly win by getting promotions, and the easiest way to do so is to optimize for short-term profits.
Yea it’s not about short or long term, it’s just about shareholders and risk management.
And honestly the system actually works pretty well. Incumbents do the thing that got them the incumbent spot, and new challengers try risky stuff to see what could beat them out. Market churn and competition fuel innovation. The cost is that incumbents occasionally get dethroned but that seems like a healthy thing overall.
Policy related to shareholders makes existing companies avoid risk in negative ways due to short term income requirements.
Established companies will error twords marginal but reliable increases in income through exploiting existing products or by buying companies after they have taken the initial risks.
This goes back to the failure of GE and is still copyed by large firms today like IBM and Cisco. Who will only grow though buying smaller companies.
The existing system does work in some ways but also stifles innovation. Basically companies are forced to sell out in order to honor short term shareholders value at the cost of long term growth.
Look at how many uprising stars are acquired, lose their culture and talent and then drop from innovation to exploiting existing customers with little product growth.
While there is a well known risk aversion problem with management in our country, much of the article's points are exactly due to short term vs long term.
It is the explore/exploit tradeoff, and the exploit phase is short term focused, always preferring small reliable increases in quarterly reports over larger long term returns.
While regulations also drove some of the older innovative groups like Bell Labs, GE, Xerox and other groups quit being innovation hubs because stock market first policies and strategies lead to a very short term focus by investors.
As a concrete example, for decades IBM would buy every analog asic compiler that would be developed just to drop development and eventually support, until the market would allow for a new startup to come around. Then IBM would buy them, rinse and repeat.
It doesn't fundamentally have to be this way, but it is, and it directly relates to short term investment and policy.
Long term there aren't new challengers, there are new acquisition targets.
IMHO ideally there would be both, innovation from startups and incumbents.
But I feel that as smaller companies have limited access to capital outside of an IPO, innovation is hampered by the death of innovative companies by acquisition and attrition that is the rule these days.
Incumbents are dethroned when they become too large and risk adverse, but typically by other massive conglomerates that are also in the short term focused exploit phase and typically not by these upstarts who are subject to aquision before they can grow to market dominance.
Putting this another way that makes it less exciting: the more important a patent is to your future survival, the higher the quality.
Or another way: multiplying the average number of forward citations by the total, yielding the absolute number of forward citations, would result in the opposite graph: massive # of forward citations for university, less for incumbent, and least for startups.
My Reinforcement Learning metaphor for equity finance driven startups is that they are the exploration process for capital seeking perfect ROI. After they find a new high margin process they exploit it to maximum financial return and (they hope) create a new paradigm that placed them as a monopolist or at worst oligopolist.
I’d argue this is precisely how major startup financiers view this - thiel explicitly says this.
So given that, and given that technology often is a key differentiator for structural change through new processes, it makes sense that evaluations of regulatory capture processes (including patents) would accrue primarily to new entrants.
Worth noting though that if you further iterate this capital/tech game, other larger monopolies/oligopolies will accrete these companies and technology too until some social tipping point is reached and the whole thing collapses.
I discuss this process as it’s root form in the third proposition of my myth of scarcity paper [1]
[1] https://kemendo.com/Myth.pdf
If you think of a three dimensional manifold as the unexplored state space of a new market, then the exploring:exploiting process maps areas of the manifold into “known” and “unknown”
It takes effort to explore the manifold in real life and the effort is measured in $/hour. Meaning whomever shows up to a new unexplored manifold with the highest potential work energy will be able to explore/search the space faster unless someone else has a more efficient search/explore or exploit mechanism.
I think this would be where “tragedy of the commons” would be relevant
Assume you are a fisherman looking for fish in the ocean
To take an exploration "step" requires work: Acquiring a boat, acquiring fishing tools, having the energy to fish, having time to explore the ocean etc...
Once you find a cache of fish, then you switch to exploit by harvesting the fish from the cache
This assumes that the fish are not currently the property of someone else, and therefore it is "commons" or a "new market" until claimed by some entity with enough power to exclude others from consumption.
The problem is, information in new markets is sparse. By definition you cannot know prior to exploring it, what the size and scope of this fish cache is. However consumption starts PRIOR to the evaluation of the cache and what an appropriate sustainable exploitation rate is because doing so would assume something else: That there is an evaluation of a global optimum for that fish cache as a common resource.
Almost no organizations have the incentive to do the above, and history shows that most consumption orgs have narrow near term goals of maximum exploitation.
So the groups with the largest amount of energy to expend exploring and then exploiting (aka how much money do you have), will by default monopolize the resource.
I don't see why the "commons" has to have anything "new". Take the example of air pollution. The air is a common resource that people have access to without cost, but it's nothing new. So the predicate that "information in new markets is sparse" doesn't apply.
>"Almost no organizations have the incentive to do the above"
I would argue that collectives, like the government, have this incentive. Hence the need for regulation.
Well, no, it's a behavioral finding.
There is nothing on the market structure stopping consolidating companies to commercialize new tech by themselves. In fact, if you take a naive view of the market, the consolidated companies are the ones with all the structure needed for it, while startups have all kinds of obstacles to get a foot on the door.
This is a well known fact, found again and again by several studies. But it is not obvious in any way.
I experience this currently at the wealth management bank I work. They were reluctant for a long time to invest into a mobile app for trading because it may cannibalise their current portfolio. They built it last year, but pricing was close to our bespoke services. The developers built a great product but when pricing is double whats on the market, then you are going nowhere.
The question "Will it cannibalise our current profits?" is asked to stop every idea. Rightfully so maybe: We make 750k revenue per employee. Money is bursting through the seams. Shame it all goes to shareholders, not employees.
[1] https://en.wikipedia.org/wiki/The_Innovator%27s_Dilemma
I wonder if this is the real objection, or really just a proxy one? As in, yes, the new thing will compete directly with the old thing, but it will easily bring in way more than it "cannibalizes". So the company will likely come ahead. However, the old thing and new thing are likely developed by distinct teams, in different business units of the company. So I imagine this question is really a proxy for the old thing team asking, "will it cannibalize our currenet jobs?". If the answer is yes, it's not surprising they'll resist.
Sony had the technology(mp3 players, phone, etc) and competence to be the first one to invent the smartphone. But I can't imagine in a risk averse culture, that a boardroom would listen to someone that would come in a suggest something that cannibalise multiple divisions in the company, in the promise that it would generate more profit, vs each divisions proposal of how to optimize their already existing profits.
And to make it even more short sighted, the idea won't go away, it will just take a little longer and then pop up in competing colours.
I think it's going to be very rare that over the life of a business every opportunity will be orthogonal to your existing business lines and not cannibalize them in some way.
Companies seem structurally incapable of cannibalizing existing business lines even if it barely impacts revenue because they cannot afford to report reduced growth for even a single quarter.
And honestly the system actually works pretty well. Incumbents do the thing that got them the incumbent spot, and new challengers try risky stuff to see what could beat them out. Market churn and competition fuel innovation. The cost is that incumbents occasionally get dethroned but that seems like a healthy thing overall.
Established companies will error twords marginal but reliable increases in income through exploiting existing products or by buying companies after they have taken the initial risks.
This goes back to the failure of GE and is still copyed by large firms today like IBM and Cisco. Who will only grow though buying smaller companies.
The existing system does work in some ways but also stifles innovation. Basically companies are forced to sell out in order to honor short term shareholders value at the cost of long term growth.
Look at how many uprising stars are acquired, lose their culture and talent and then drop from innovation to exploiting existing customers with little product growth.
While there is a well known risk aversion problem with management in our country, much of the article's points are exactly due to short term vs long term.
It is the explore/exploit tradeoff, and the exploit phase is short term focused, always preferring small reliable increases in quarterly reports over larger long term returns.
While regulations also drove some of the older innovative groups like Bell Labs, GE, Xerox and other groups quit being innovation hubs because stock market first policies and strategies lead to a very short term focus by investors.
As a concrete example, for decades IBM would buy every analog asic compiler that would be developed just to drop development and eventually support, until the market would allow for a new startup to come around. Then IBM would buy them, rinse and repeat.
It doesn't fundamentally have to be this way, but it is, and it directly relates to short term investment and policy.
Long term there aren't new challengers, there are new acquisition targets.
IMHO ideally there would be both, innovation from startups and incumbents.
But I feel that as smaller companies have limited access to capital outside of an IPO, innovation is hampered by the death of innovative companies by acquisition and attrition that is the rule these days.
Incumbents are dethroned when they become too large and risk adverse, but typically by other massive conglomerates that are also in the short term focused exploit phase and typically not by these upstarts who are subject to aquision before they can grow to market dominance.
My immediate thought is: "If we don't cannibalise their current portfolio, somebody else will" and use that as the argument to carry on on good ideas.
Why do you think that has not been the natural inclination?
Or another way: multiplying the average number of forward citations by the total, yielding the absolute number of forward citations, would result in the opposite graph: massive # of forward citations for university, less for incumbent, and least for startups.
It is almost a proxy for patent cost.