Google Technologies: Staying Ahead of the Trends (Essay Sample)

Google Technologies
Your Name Goes Here
American Public University System
Google Technologies
Google is a multinational American Corporation that is known for its specialization in the development of internet-based products and services. However, the organization’s research and development department explores various technological aspects through research, aiming at influencing product creation in different industrial sectors. The goal of the research efforts made is to unearth discoveries that impact lives across the globe, with the approach applied being one that advocates for the sharing of both research findings and tools to fuel progress in the associated sectors. As such, Google researchers publish their results in academic journals regularly and release their projects as open sources to allow other industrial players, in the fields explored, to access the fundamentals thereof and expand their knowledge base on the subject matter (Google Research, 2020). Thus, Google’s research and development endeavors are geared towards boosting both the organization’s achievements as well as facilitating the betterment of different industrial sectors.
The paper will explore five Google technologies, including self-driving automobiles, Deep Graph Infomax, remote diabetic retinopathy, Open Learning Initiative, and Probabilistic Object Detection. Google’s exploration of self-driving cars began as a project called Waymo, which expanded and became a full company in December 2016 (Gannes, 2014). Secondly, in its machine intelligence work, Google is exploring different aspects of machine learning including classical algorithms, as evidenced by the Deep Graph Infomax (DGI) research (Google Research, 2020a). Thirdly, in health and bioscience, the organization is looking to exploit the unique potential to enhance people's lives by seeking a greater understanding of scientific aspects that are crucial to diagnosing diseases using advanced techniques such as remote diabetic retinopathy (Google Research, 2020b). Fourthly, Google’s research on education innovation focuses on large scale publication of online learning, which explores all aspects of tech-based teaching such as the open learning initiative (OLI) (Google Research, 2020c). Lastly, in robotics, Google seeks to foster collaboration among researchers in the stated industry and machine learning to facilitate the creation of study environments that exploit both simulated and real robot systems, as exemplified by the probabilistic object detection (Google Research, 2020d). The research objectives outlined above align with Google’s mission which is to organize data on a global scale and avail it to all. Additionally, the research goals thereof match the organization’s vision of availing comprehensive information at a single click, especially considering the corporation’s research orientation towards the maximization of efficiency (Thompson, 2019). Hence, through the research and development projects that Google undertakes, it pursues the fulfillment of both its vision and mission.
Self-Driving Automobile
According to Gannes (2014), Google built is the first self-driving car, which was a two-seater vehicle, and unveiled its creation in the same year. Unlike the previously modified models that Google had worked with, the prototype that was introduced in 2014 lacked a steering wheel, pedal, and brake pads. Additionally, among the other parts of a standard car that were missing in the new type automobile were the glove compartment, stereo, and the backseat. Instead, the vehicle was endowed with multiple sensors and software to guide the self-driving function, which was created by Google for use in Lexus SUVs and Toyota Priuses. As such, the installed software had been conditioned to meet highway and city streets driving needs over the preceding five years.
Among the challenges that Google could face when implementing the self-driving automobile technology is the concern that would arise about the safety implications and legal liability thereof. Even though the robotic cars are likely to be more practical and accurate on the roads compared to error-prone human drivers, it is an indisputable fact that engineered systems cannot be flawless. Moreover, in the case that the self-driven automobile fails and an accident occurs, the determination of who should be held responsible could elicit a complicated and non-yielding fault-finding process. Therefore, to overcome the stated challenges, Google must ensure that it meets the predetermined safety benchmark, which will require that the self-driven automobiles cover more than 750,000 miles without causing accidents to confirm that they are at least 99% crash-safe compared to ordinary cars. Furthermore, Google must come up with a liability system that can be used in case of accidents. For instance, the organization can choose to employ a no-fault liability system that allows damage recovery by victims from their car insurers. In a different case, Google can pose a legal argument establishing the irrebuttable assumption that owners have control over autonomous vehicles, which applies regardless of the delegation of control to the car (Bailey, 2014). The definition of a liability system and provision of proof guaranteeing that the self-driving vehicles meet and surpass standard safety criteria will allow Google to introduce the technology with minimal resistance.
Self-driving cars are likely to impact the environment positively because it has been proven that they minimize energy consumption by up to 90%. The latter will translate into lesser exploitation of finite environmental resources as well as the reduction of the greenhouse gases that are emitted into the atmosphere, consequently resulting in a decrease in global warming rates. However, the efficiency of autonomous cars would make them a highly convenient mode of transportation that would be increasingly used, resulting in heightened energy consumption by up to 200% (Chang, 2018). Such adversities could be mitigated by ensuring that limitations on the use of autonomous cars are implemented before the mass rollout of the stated technology. Additionally, the security and privacy concerns that would be associated with autonomous vehicles would be those that are linked to hacking. The latter is a cybercrime that exploits weaknesses in computer systems and can be used by malicious third parties to alter their functionality (Pavlik, 2017). For instance, if a self-driving car were hacked, its owner would lose control of the vehicle, a factor that could make it impossible for them to safeguard their lives and those of other road users. Therefore, to address such concerns, Google would need to equip the self-driven automobiles with well-fortified security systems that can detect and protect against malicious activity. Moreover, the introduction of autonomous autos will make the management of information systems more efficient because tracking systems will only need to keep tabs on the cars and not their drivers. As such, managers dealing with transportation will have an easier time assessing accountability because they will only require to access the vehicles tracking systems for the information needed. Regardless of the cons that are associated with the adoption of autonomous vehicles, there is a moral obligation to embrace the technology because it will save human lives and enhance well-being by lowering the healthcare costs incurred when accidents occur as a result of human error.
Deep Graph Infomax
The Deep Graph Infomax (DGI) technology characterizes one of the research areas that Google is pursuing in machine learning. DGI uses a generalized approach to represent learning nodes in cases that involve data that is graph-structured. It maximizes shared information between correlated complex graph summaries and patch representations, which summarize data in the sub-graphs that are around the nodes in focus. Hence, the stated images can be reused to fulfill downstream learning objectives. Unlike most prior approaches to machine learning, such as those represented by GCNs, DGI can be applied in both inductive and transductive learning apparatus because it does not rely on random objectives (Veličković et al., 2018). Below is a representation of the DGI algorithm.
Among the issues that Google could face when introducing the DGI technology is the fact that the graph neural networks (GNNs) used for homogeneous graphs may be incapable of handling relations involving different semantics, as is the case with heterogeneous graphs. Thus, to overcome the stated challenge, the organization should consider incorporating a meta-path that takes the same form as the neural network model used for unsupervised heterogeneous graphs. The latter would result in the creation of the Heterogeneous Deep Graph Infomax (HDGI), which would be more efficient compared to the DGI (Ren et al., 2019). Additionally, since DGI represents machine learning technology, which can explore data and learn for itself, Google would face the challenge of proving that the reprogramming thereof would have minimal negative implications. Faria (2018) explains that the engineering of a safe system is associated with validation uncertainties, and in the case of machine learning techniques such as DGI, the levels of unpredictability are heightened because it is impossible to determine whether the algorithm used can predict correctly. Therefore, to curb the stated concern, Google must ensure that the DGI data set training is executed thoroughly, to ensure that the model selected is the most efficient and the optimization strategy applied minimizes the unpredictability of the algorithm.
DGI can impact the environment positive...