リテイシエフ デイミトリ

The phenomenon of young people (under 30) starting or working in ventures is common in Silicon Valley but rare in Japan. Avoiding cultural attributions upon which many international comparisons of entrepreneurship are truncated, we apply strategic behavior theory to uncover a rational-choice basis for this phenomenon. We identify individual and organizational players, consider their strategies, and compare equilibria in the institutional context of Japan and the US. We pay close attention to how competition in Japanese educational, labor, and marriage markets differs from such competition in the US to identify factors which raise the career attraction of big firms and thereby fuel adverse selection that hurts ventures. Our conclusions challenge the stereotype that the founders and employees of Silicon Valley ventures are heroic risk-takers whereas the Japanese are much more risk-averse.

The paper clarifies how sunk costs can lead a rational incumbent to innovate less than an entrant. It also demonstrates that competition among incumbents yields less adoption of new and more efficient production technology than competition which includes entrants. The results suggest that policy promoting adoption of next-generation production technology should distinguish among firms based on their sunk costs in current-generation technology and encourage entry by industry outsiders such as startups or firms from other industries.

Competition among profit-seeking firms in an oligopolistic industry inherently generates incentives for firms to commit to maximize a performance metric other than profit.We briefly review the underlying theory,analyze its ramifications in a Cournot duopoly,and consider feasibility constraints from the perspective of strategic management.

Why are humans so vulnerable to pain in interpersonal relations and can so easily hurt others physically and emotionally? We theoretically examine whether being offensively strong but defensively weak can evolve as a strategic trait that fosters cooperation. We study a population comprised of "thick-skinned" and "thin-skinned" agents by using an indirect evolution model that combines rational choice in strategic interactions with evolutionary selection across generations. We find that (a) the relatively vulnerable and cooperative thin-skins cannot evolve under purely random matching, (b) with some assortment thin-skins evolve and can take over the entire population, (c) vulnerability to greater pain makes it easier for thin-skins to evolve, and (d) proximate pain which merely feels bad but does not lower fitness helps thin-skins evolve even more than pain which accurately reflects fitness consequences. We draw contrast with the Hawk-Dove model and identify several ways in which rationality hinders the evolution of the relatively vulnerable and peaceful type of agent.

An outside inventor of a new production process seeks to license it to Cournot duopolists which have unequal ex ante costs. Distinguishing "leading-edge" innovations (new cost below both firms' costs) from "catch-up" innovations (new cost between the two firms' costs), we compare the equilibria of two license-selling mechanisms: exclusive license auction and non-exclusive price-setting. In contrast to the often-studied case of an innovation that reduces the cost of any licensee by the same amount, we show that licensing of a new process may attenuate the ex ante cost asymmetry, allow the inefficient firm to leapfrog its competitor, and raise the licensee's net profits.

A parsimonious model is used to explore the risk-bearing decision under a payoff structure that emphasizes relative performance. Equilibrium betting amounts are derived for players who start with unequal endowments and face a lottery that offers either a positive or negative expected return. If the lottery offers negative expected return, disadvantaged players are willing to risk a portion of their endowment, and this induces advantaged players to also gamble, defensively. Although there are equilibria in which the advantaged preemptively gamble more than the disadvantaged, in the robust equilibrium it is the disadvantaged who make the larger bets. If the lottery offers positive expected return, there are equilibria in which the advantaged invest less than the disadvantaged, but full investment by all players is a more robust equilibrium.

A recent trend in automatic speech recognition systems is the use of continuous mixture-density hidden Markov models (HMM's), Despite the good recognition performance that these systems achieve on average in large vocabulary applications, there is a large variability in performance across speakers, Performance degrades dramatically when the user is radically different from the training population, A popular technique that can improve the performance and robustness of a speech recognition system is adapting speech models to the speaker, and more generally to the channel and the task, In continuous mixture-density HMM's the number of component densities is typically very large, and it may not be feasible to acquire a sufficient amount of adaptation data for robust maximum-likelihood estimates, To solve this problem, we propose a constrained estimation technique for Gaussian mixture densities, The algorithm is evaluated on the large-vocabulary Wall Street Journal corpus for both native and nonnative speakers of American English, For nonnative speakers, the recognition error rate is approximately halved with only a small amount of adaptation data, and it approaches the speaker-independent accuracy achieved for native speakers, For native speakers, the recognition performance after adaptation improves to the accuracy of speaker-dependent systems that use six times as much training data.

A statistical technique for vector quantizer (VQ) prototype adaptation, based on tied-mixture continuous-parameter HMM's, is derived and evaluated on the basis of experimental evidence. Performance on difficult adaptation tasks indicates that VQ-prototype adaptation via tied-mixture HMM's constitutes a useful mechanism for speaker adaptation, particularly when there are substantial channel differences or when there is a large mismatch between reference and target speaker characteristics.