热门搜索：

2024测试生成式人工智能：降低风险和最大化机会报告（英）

信息技术2024-07-06-Applause郭***

AI智能总结

Generative AI: Overview, Applications, Trends, and Opportunities

Basics of Generative AI

Generative AI (Gen AI) is a technology that creates content based on user requests through a process known as "prompting." This technology synthesizes, summarizes, or generates new text, images, video, and audio content by learning from vast amounts of human-generated data. Unlike search engines, which retrieve information in its original format, Gen AI uses large language models (LLMs), a type of neural network, to analyze data and provide value-added responses.

Understanding Generative AI

Gen AI encompasses applications such as ChatGPT, Gemini (Google), and Copilot (Microsoft), among others. These applications utilize LLMs and diffusion models to interpret and generate text and can also handle image, audio, and content generation. The term "multimodal" describes LLMs capable of processing multiple content formats simultaneously.

Differences from Traditional AI Technologies

Gen AI diverges from machine learning (ML) and natural language processing (NLP) in its primary function. ML focuses on pattern recognition and data analysis, whereas Gen AI specializes in content creation and transformation. NLP is more suited for interpreting human speech or text and providing predefined responses, lacking the versatility of Gen AI in handling complex information-based queries.

Current Trends and Use Cases

Today, Gen AI is predominantly used for consumer entertainment and day-to-day task assistance, as well as for boosting enterprise productivity. Among the most popular use cases are text-driven inputs and outputs. Internal solutions are favored by organizations for testing purposes, but external solutions are expected to yield greater impact, especially in customer support and core product integration.

Future Outlook and Opportunities

The adoption of Gen AI is characterized by three primary use cases. Companies are increasingly considering integrating Gen AI into their product offerings, either by creating new AI-powered products or enhancing existing ones. The trend towards narrow domain solutions, focusing on specific tasks, is gaining prominence over broad-domain solutions. The future of Gen AI promises hyper-automation, hyper-creation, and hyper-personalization, offering significant opportunities for innovation and efficiency across industries.

Key Takeaways

Gen AI leverages LLMs to automate tasks requiring human-level intelligence.
Applications range from text generation to multimedia content creation.
Integration into existing products is a preferred strategy over developing new AI-only solutions.
Narrow domain solutions are emerging as a fast-growing segment in Gen AI.
Opportunities lie in hyper-automation, hyper-creation, and hyper-personalization.

This overview highlights the foundational aspects, applications, trends, and future possibilities of Generative AI, emphasizing its transformative potential in enhancing productivity and personalization across various sectors.

EBOOK TESTINGGENERATIVEAI:MITIGATINGRISKSANDMAXIMIZINGOPPORTUNITIES UNDSEERCSTAIONNDING 1 GETNITELREAHTIEVREEAI TheBasicsofGenerativeAI Anoveltechnologythatcreatescontentbasedonuserrequestsknownasprompts,GenerativeAI(GenAI)applicationscanproducenewtext,images,videoandaudiobysynthesizing,summarizing,orgeneratingcontent.WhileusersinteractwithGenAImuchliketheydoasearchengine,thewaythetwotechnologiesgenerateresponsesisverydifferent. Whereassearchenginesretrieveinformationintheformatinwhichitisstored,GenAIanalyzeshugeamountsofhuman-generateddata,knownastrainingdata,tolearnhowtorespondtouserrequestsinawaythatprovidesvalue.Theanalysisiscarriedoutbyalargelanguagemodel(LLM),atypeofneuralnetwork.Overtime,humansmustcontinue toprovideLLMswithfreshtrainingdatatofine-tunethemodelandkeeptheinformationitlearnsfromuptodate. GenAIcanautomateandenhanceanynumberofbusinesstasks,rangingfromdraftingemailsandgeneratingfinancialreportstocreatingmarketingcontentandanalyzingcustomerinteractions—andusecasesareexpandingdaily.Today,thetechnologyislargelyusedtoimproveproductivityandpersonalizeuserexperiences. PopularGenAIapplicationsincludeChatGPT(OpenAI),Gemini(Google)andCopilot(Microsoft).Whileeachapplicationcurrentlymainlyofferstext-basedinputs,voiceinteractionisontherise.Midjourney,StableDiffusionandDALL-Earepopularapplicationsintheimagegenerationcategory. TestingGenerativeAI:MitigatingRisksandMaximizingOpportunities Keytermsexplained Largelanguagemodels(LLMs) AIfoundationmodels,includinglargelanguagemodels(LLMs)anddiffusionmodels,powerthevastmajorityofGenAIapplications.Theseinnovationsrepresentarevolutionarydevelopmentinthehistoryofhumanitybecausetheyautomatetasksthathavehistoricallyreliedentirelyonhumaneffort,suchastextgeneration,summarizationandanalysis.JustastheIndustrialRevolutionautomatedlabor,thegenerativeAIrevolutionisautomatingintelligence.Infact,itislikelythatthecurrenterawillbereferredtoastheintelligencerevolutionbyfuturegenerations. LLMsarewidelyusedinapplicationsthatinterpret,transformorgeneratetext.However,theyarealsoemployedinusecasesthatrequireimage,audio,andcontentgeneration.Inaddition,AImodelsthatgeneratevisualandaudiocontentoftenemployLLMstointerprettheuser’sinputrequestandaresometimesusedtoevaluatepromptinputsoroutputstorespondtouserquestions. ItisworthclarifyingthatanLLMisnotanapplication.Itisatechnologythatsupportsanapplication.Forexample,ChatGPTisanapplicationpoweredbyamodelcalledGPT-4o.Claude,atext-basedGenAIapplication,ispoweredbytheLLMsHaiku,SonnetandOpus.MetaAI,whichformspartoftheFacebook,InstagramandWhatsAppapplications,isbasedontheLLMLama3. MultimodalLLMs MultimodalisatermemployedtodescribeLLMsthatcanprocessinformationprovidedbyusersinmultiplecontent‘modes,’suchastext,images,diagramsandvideos.Theycanalsogeneratenewcontentinthesemodes.Theleadingmultimodalmodelscananalyzeandgeneratevariouscontentmodesaspartofasingleprompt. HowGenerativeAIDiffersfromTraditionalAI GenerativeAIisoftenconfusedwithtwootherAItechnologies:traditionalmachinelearningandnaturallanguageprocessing. »Machinelearning(ML)istypicallyemployedtoanalyzelargequantitiesofdataandidentifypatterns,anomaliesorhiddeninsights.WhileGenAIalsoemploysML,itismainlyusedtogenerateortransformcontent,ratherthananalyzeit.UnliketraditionalML,whichoftenfunctionsasynchronouslyfordataanalysis,GenAItypicallyoperatesasaruntimetechnology. »Naturallanguageprocessing(NLP)ismostcommonlyusedinvoiceassistantsandchatbots.Itisusedtointerprethumanspeechortext(suchaswhenauserverballyasksavoiceassistantlikeAmazonAlexaaquestion)andthenmatchtheirquestionwithapredefinedanswer.WhileNLPcanprocessunboundeduserinput,itcanonlyprovidepre-determinedoutputs.ThismakesitlessversatilethanGenAIforrespondingtoinformation-relateduserintents. TestingGenerativeAI:MitigatingRisksandMaximizingOpportunities AdamCheyer,thecreatorofthetechnologybehindApple’sSiri,saidinapodcastthatNLP-basedsolutionslikeSiriwereoptimizedas‘doing’assistants,likeplayingmusicoropeningauser’scalendar.Bycontrast,manygenerativeAIapplicationsare‘knowing’assistants. Theyexcelatinformation-basedqueriesandcreativerequestsbutarenotasadvancedinconsistentlyexecutingproceduralrequests. ExamplesofNLP,MachineLearningandGenAI NaturalLearningProcessing »Alexa »GoogleAssistant »Siri »BankofAmerica’sErica »Mostwebsitechatbots TraditionalMachineLearning »Proprietarybusinessanalytics »Forecastingmodels »Nottypicallyusedbyconsumersbutsometimesexposedthroughalgorithms GenerativeAI »ChatGPT »MicrosoftCopilot »GoogleGemini »MetaAI »Midjourney »StableDiffusion SEC2TIONGENERTAITTLIVEEHAEIRTERENDS SeveralcurrenttrendscharacterizeGenAIadoptiontodayandwhereit'sheaded.Theyareexploredbelow. Threeusecasesarebyfarthemostpopular TounderstandGenAI,youneedtoconsideritsapplicationforconsumersandenterprisesalike.Consumersareinterestedinusingitforentertainmentpurposesandforassistancewithday-to-daytasks,whileenterprisesarefocusedonleveragingittobooststaff productivity.Allotherusecasescomeadistantsecondintermsofpriority. Text-driveninputsandoutputsarel

点击免费查看完整报告