生成式人工智能将加速金融公司的竞争力

Google Cloud Survey ShowsBanks Ready to Embrace AI Will financial institutions achieve the transformativeopportunities open to them if they embrace Generative AI(Gen AI) to automate business processes, energise data,and fortify decision-making? According to a recent surveyby Google Cloud1the answer is a resounding “YES” assome 92% of the banking executives questioned not onlysaid there is high demand for Gen AI within the bankingsector, but 95% believed it has the potential to transformthe industry. Bank Support forAI Booms SAY AI WILL TRANSFORMTHE INDUSTRY At its most fundamental Gen AI is also expected to makebanks and insurers more competitive by enriching theway they interact with customers to build loyalty and growrevenues. EXPECT UP TO 50% BUDGETRISE FOR AI Bankers are not only beginning to understand thisopportunity, but it looks like they are putting their moneywhere their mouths are. Some 40 percent of respondentssaid Gen AI deployment will require a 26%-50% increasein their IT budgets over the next 1-2 years. While we knowthat banks and insurers are already concerned about ITcosts, it appears that extra money is expected to belargely self-financed from the increased ROI and reducedTCO that Gen AI will deliver. SEE 60%+ COST SAVINGSFROM AI WILL SEEK PARTNERS FORAI STRATEGIES Some 49% of the banking executives surveyed said thetop benefit Gen AI can bring is increased operationalefficiency and cost savings. Specifically, 38% of them saidthat Gen AI will deliver 61%-80% in cost savings over thenext five years. This was reinforced by internal research into Gen AI-led productivity benefits conducted bySoftServe data science teams. These showed a 31% reduction in the time needed to bring newcapabilities to market and a 45% increase in output when deploying Gen AI. The study engagedmore than 1,000 technicians to run 1,500 experiments on over 100 use cases in seven countries. Banks and insurers also acknowledge that they willnot be able to go it alone if they are to achieve theoptimum benefits of Gen AI, as part of that budget isbeing earmarked for third-party AI partnerships.Interestingly, an overwhelming 93% of the bankerssaid their organizations plan to partner with AItechnology providers within the next 6-12 months tosupport their Gen AI strategies. From the start, Gen AI enables users to quicklygenerate new content based on a variety of inputs.These inputs can include text, images, sounds,animation, code, or other types of data. It usesneural networks to identify patterns and structureswithin the existing datasets to generate that content. In essence, Gen AI automates different learningapproaches. These enable organisations to quicklydeploy large amounts of data to create foundationmodels that can be used as a base for AI systems. This paper aims to show where Gen AI can make thebiggest differences for finance firms and how it canbe deployed and integrated to enhance existing ITinfrastructure. But it is also designed to show howwe work with expert technology partners like GoogleCloud to enrich the discussions around the benefits(and risks) of using Gen AI and to stimulate betterinformed debate within banking boardrooms. These foundation models, says Google Cloud, canmulti-task and perform out-of-the-box tasks,including summarization, Q&A, classification, andmuch more. Plus, with minimal training required,foundation models can be adapted for targeted usecases with very little example data. It is one of the reasons we recently expanded our“preferred partner” relationship with Google Cloudto offer the Google Cloud Gen AI Use Case DiscoveryWorkshop. This is a service from SoftServe thatprovides clients with the tools to pursue GoogleCloud Gen AI technologies and increase their valueproposition in the market. But more on that later. Gen AI works by using an ML model to learn thepatters and relationships in a dataset of humancreated content. It then uses the learned patterns tocreate new content. The most common way to traina Gen AI model is to use unsupervised learning—the model is given a set of human created contentand corresponding labels. It then learns to generatecontent that is similar to the human created contentand labelled with the same labels—but in a fractionof the time. Gen AI isn't just a new technologybuzzword—it has the potential torevolutionize the way we live, work,bank, and invest," Ultimately, Gen AI becomes a powerful tool thatstreamlines the workflow of creatives, engineers,researchers, scientists, and more, with use casesthat span all industries and individuals. And it can beeasier to deploy than many expect. Google Cloud’s Vertex AI offering can interact with,customize, and embed foundation models into bankapplications, without requiring ML experience. Itenables users to access foundation models on ModelGarden, tune models via a simple UI on GenerativeAI Studio or use models in a data science notebook.Its Vertex AI Search and Conversation tools offerdevelopers th