{ "docId": "019dd923-5ca1-7489-b638-8d31fa857bad", "docSlug": "01d275675630cbbe", "documentTitle": "How nine digital frontrunners can lead on AI in Europe", "authorId": "McKinsey", "authorName": "McKinsey", "documentKindSlug": "consulting-deck", "documentKindLabel": "Consulting deck", "sourceTypeSlug": "strategy_consulting", "sourceTypeLabel": "Strategy consulting", "presentationDate": null, "orientation": "landscape", "aspectRatio": 1.445, "pageNumber": 9, "pageCount": 66, "prevPage": 8, "nextPage": 10, "slideType": "industry_trends", "function": "establish_context", "density": "overcrowded", "nDataPoints": 5, "notes": "The slide provides a narrative summary of AI progress in specific domains with supporting data points.", "elementsJson": [ "paragraph", "footnote" ], "metadataConfidence": 0.95, "imagePath": null, "slideHref": "/slides/019dd923-5ca1-7489-b638-8d31fa857bad/9", "deckHref": "/decks/019dd923-5ca1-7489-b638-8d31fa857bad", "deckJsonHref": "/decks/019dd923-5ca1-7489-b638-8d31fa857bad.json", "deckAnchorHref": "/decks/019dd923-5ca1-7489-b638-8d31fa857bad#slide-9", "components": [ { "bbox": { "h": 0.1, "w": 0.38, "x": 0.1, "y": 0.39 }, "kind": "paragraph", "text": "In most environments, AI's abilities remain inferior to those of humans. For example, an AI-powered car has trouble interpreting the movements of pedestrians. However, in some cases, it surpasses humans, and its abilities are accelerating as research and funding increase (Exhibit 3). There has been particularly fast growth in three areas:", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "26e1b86a-aba5-45a1-bc54-1a79ec3a7748", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.35, "w": 0.38, "x": 0.1, "y": 0.5 }, "kind": "paragraph", "text": "Image recognition. One of the most common uses of AI is image recognition. The modern smartphone user – always connected to the internet and equipped with a high-quality camera – produces and consumes large quantities of image and video content every day. AI image recognition technology surpassed the human accuracy level of 95 percent in 2015 and can be used to categorize, edit, and parse this data. Applications building on this technology are being rolled out at scale on social media, for instance, where facial recognition is used for tagging. Applications extend beyond our browsers into the physical world, despite regulatory uncertainty. For example, Alibaba's grocery venture Freshippo uses it to verify payments, and Shanghai's Hongqiao Airport employs a facial-recognition-based check-in system.", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "7ff81c7c-a4bc-49d6-9347-d4cfd9aa3129", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.15, "w": 0.38, "x": 0.1, "y": 0.23 }, "kind": "paragraph", "text": "Though the concept of AI is half a century old, interest in the technology has grown rapidly over the past decade. This has been driven by increased data availability and computing power, the declining cost of data processing, and new advanced mathematics, allowing for significantly lower costs of prediction and analytical capacity. These developments mean organizations can use the technology's ability to interpret large volumes of data, enabling them to perform more complex and targeted calculations.", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "81dd1337-6a7c-497b-ae95-ec1354ba9a52", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.18, "w": 0.38, "x": 0.52, "y": 0.63 }, "kind": "paragraph", "text": "In some areas, AI is moving beyond image, speech, and text to abilities considered even more uniquely human. A burgeoning field of study is affective computing, which seeks to equip AI with the ability to recognize, interpret, and replicate human emotions. A 2018 Ohio State University experiment found that AI could recognize some emotions more accurately than humans by going beyond interpretations of facial expressions alone, to include skin color, body posture, and cultural context.", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "9c9fd79d-1523-4ebd-b274-947dd7292eda", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.18, "w": 0.38, "x": 0.52, "y": 0.44 }, "kind": "paragraph", "text": "Natural language processing. There is a vast amount of information in text. Natural language processing (NLP) can contribute to recognition, understanding and generation of language and has been shown to exceeded the average human score in reading comprehension tests of 86 percent in 2018. NLP can be combined with speech recognition to process the meaning from speech or with image recognition to analyze handwriting. NLP AI can help businesses make smarter decisions, governments deliver better services, or simply make peoples' lives easier!", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "9d573c7d-1aa7-43d6-a410-3949a33ffe93", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.25, "w": 0.38, "x": 0.52, "y": 0.18 }, "kind": "paragraph", "text": "Speech recognition. AI can also be used to translate spoken words into text. AI speech recognition surpassed the human accuracy level of 95 percent in 2017 and some of the world's leading tech companies use it on devices such as Google Home, Amazon Echo, and Siri. While the technology is still in its infancy, it has significant potential. Research firm Gartner predicts that 30 percent of web browsing will be without a screen in 2020. Another use case is for emergency services, which incorporate algorithms to identify callers' symptoms by listening to a person's breathing, voice, and words.", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "db4da5cd-e687-45e1-9be4-95b23cde8ee8", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.08, "w": 0.5, "x": 0.1, "y": 0.88 }, "kind": "source-note", "text": "1 Gartner research, 2017\n2 The conversion from e.g., handwritten to machine encoded text is known as Optical Character Recognition (OCR)\n3 Web page of The Ohio State University\nSource: Stanford University HAI, 2019; ImageNet; Kleiner Perkins, 2018; Stanford Questions Answering Dataset; CodaLab", "attrs": null, "subkind": null, "toolName": null, "toolSlug": null, "confidence": null, "componentId": "b89377a5-a909-4f06-beb2-d5e5eba1cc5d", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.03, "w": 0.35, "x": 0.1, "y": 0.18 }, "kind": "title", "text": "AI has started to surpass human accuracy", "attrs": null, "subkind": "headline", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "0c613ded-10fc-461e-8e23-570375ad78c9", "frameworkName": null, "frameworkSlug": null } ], "metrics": [], "tools": [], "frameworks": [], "arcBeats": [], "loops": [ { "to": 20, "from": 4, "name": "Golden Circle", "slug": "11-golden-circle", "bestFor": "Visionary leadership, brand positioning, mission statements", "matchId": "e7fed04a-464b-4a94-9e13-0ff55e265f76", "evidence": "The document uses the golden circle framework to discuss the why, what, and how of AI for the DF9", "position": 0, "objective": "Why AI is important for the DF9", "structure": "The Why (Belief) -> The How (Process) -> The What (Result)", "confidence": 0.8, "description": "Invert the typical pitch by starting with why you exist, rather than what you do" } ], "imagePathAlt": null, "thumbSrc": null, "thumbSrcAlt": null, "locked": true }