1. Research
21. März 2023
Artificial Intelligence arguably came to life for the general population with the release of the accessible chatbot ChatGPT three months ago, but the technology and opportunities likely remain a mystery to many. Following March 14, 2023 release of GPT-4, an update to the technology underlying ChatGPT, we publish a new Chartbook on Artificial Intelligence and address the five Ws: Why, What, Who, When, and Where. [mehr]
Deutsche Bank Research Why, What, Who, When, Where? DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MCI (P) 051/04/2022. UNTIL 19th MARCH 2022 INCOMPLETE DISCLOSURE INFORMATION MAY HAVE BEEN DISPLAYED, PLEASE SEE APPENDIX 1 FOR FURTHER DETAILS. March 2023 Marion Laboure, Ph.D - Global Research marion.laboure@db.com Cassidy Ainsworth - Grace cassidy.ainsworth-grace@db.com Adrian Cox adrian.cox@db.com #PositiveImpact AI & The Five Ws: AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 2 Artificial Intelligence suddenly came to life for the general population with the release of the accessible chatbot ChatGPT three months ago, but the technology and opportunities are still a mystery to many. Following the release of GPT - 4 on March 14, an update to the technology underlying ChatGPT, we publish a new Chartbook on Artificial Intelligence and address the five Ws: Why, What, Who, When, and Where. Part 1 - What is artificial intelligence is. We take a look at "what" and "why" of artificial intelligence: its genesis, the different types of AI and the skills someone needs to work in the field . Part 2 - Today's artificial intelligence landscape. We take a look at today's artificial intelligence landscape, including the rapid growth of its capabilities and investment into the field. Part 3 - What the opportunities for corporates are. We show how corporates are becoming increasingly interested in the development of artificial intelligence, and how smaller corporates may be swept up in an upspring of innovation after OpenAI released its API. Part 4 - Capabilities: Where we are today and where we are moving to . We illustrate how artificial intelligence has become faster, more affordable, but has placed increasing demands on technological capabilities. We examine the sustainability of artificial intelligence technology, before analysing the nature of regulation on the sector. …All in an easy - to - read PowerPoint format. Chartbook: AI & The Five Ws: Why, What, Who, When, Where? What Where When Who Why AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 3 Open AI's ChatGPT and GPT - 4 : A quick overview What are ChatGPT and GPT - 4? - ChatGPT is a chat interface built by OpenAI that allows the public to interact with its AI model, GPT - 3.5. It was released on Nov. 30, 2022. - GPT - 4 is OpenAI's latest AI model, released on March 14, 2023, and the update to GPT - 3.5. - It is available on subscribers to ChatGPT Plus and as an Application Programming Interface (API) for developers to build applications and services. - The models are "trained" using huge volumes of data and generate text and code based on the relative probability of what should come next What is the difference between the two models? - GPT - 4 is multi - modal, using deep learning to digest not just text but also images, and to respond with text. That includes suggesting a recipe from a photo of ingredients and correctly interpreting the humour in a photo of a phone with an old - fashioned plug. - It outperforms ChatGPT on creative and reasoning metrics, according to OpenAI . - It is 82% less likely to respond to requests for disallowed content and 40% more likely to produce factual responses. It can take in and generate around 25,000 words of text, around eight times more than ChatGPT. - GPT - 4 scores above ChatGPT in common tests, reaching the 90 th percentile in the Uniform Bar Exam, compared with the 10 th . Where do they fit into the market? - After the launch of ChatGPT, Microsoft stepped up its financial and computing power support for OpenAI . A wave of other AI model announcements is taking place, with Google saying earlier on March 14 it will open new generative AI tools to some clients and Google - backed Anthropic launching its Claude chatbot. Microsoft is expected to give details on how it will roll out AI features for Office software later in the week. Capabilities of Large Language Models - Retrieve and format information - Classify sentiment, common topics and indicators of risk from data - Engage in Q&A - Analyse and summarise text - Translate text - Generate text, computer code, images, music etc Limitations - Social biases, based on the data they "learn" from - "Hallucinations" in the form of invented facts - Data limitations, eg ChatGPT/GPT - 4 data set largely stops in 2021 Metaverse & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 4 INTEREST: The number of AI public projects and annual patent fillings has exploded since 2015 Number of public AI projects 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Source: Deutsche Bank, OECD.AI. Annual number of patent filings for AI technologies 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | - Google searches for "artificial intelligence" skyrocketed in November 2022, following the release of OpenAI's ChatGPT . INTEREST: P eople have recently increased searches on these topics 5 Source: Deutsche Bank, Google Trends. Updated 8 March 2023. Google trends, worldwide web searches 0 20 40 60 80 100 Jan 2022 Mar 2022 May 2022 Jul 2022 Sep 2022 Nov 2022 Jan 2023 ChatGPT Artificial intelligence OpenAI Jan 2023: Microsoft announces new investment in OpenAI . Feb 2023: Google releases its AI tech BARD. Sep 2022: DALLE - 2 released by OpenAI. Nov 2022: ChatGPT launched by OpenAI. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 1. What is AI " " 6 "We're at the beginning of a golden age of AI. Recent advancements have already led to inventions that previously lived in the realm of science fiction - and we've only scratched the surface of what's possible." - Jeff Bezos, founder of Amazon AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | HISTORY : The genesis of artificial intelligence 1943 1950 1956 1966 1972 1974 - 1980 1980 1987 - 1993 1997 1996 2000 2002 2011 2016 2015 2015 2017 2022 Artificial neurons Turing Test First chatbot: ELIZA First intelligent robot: WABOT - 1 First AI winter AI in homes: Roomba IBM's Watson wins quiz show Modern Google Amazon Echo 7 Dartmouth Conference Massive Internet Usage Expert system Second AI winter IBM Deep Blue OpenAI founded ChatGPT released Google Lens - The concept of artificially intelligent beings and machines has been around since the 19 th century, with Mary Shelley's 1818 Frankenstein, Samuel Butler's 1972 Erewhon and the Maschinenmensch in Fritz Lang's Metropolis in the early 20 th century . - AI in the field of science came to life with Alan Turing's 1950 paper on Computing Machinery and Intelligence, solidifying it s place in the history of science with the 1956 Dartmouth Conference, where the term 'artificial intelligence' was coined. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 8 Source: Deutsche Bank, Deloitte. Artificial intelligence: Ability to sense, reason, engage and learn Includes Natural Language Processing, voice recognition and robotics AI DEFINITION: Computer systems that can perform tasks that require advanced, often human - like, intelligence • OECD definition: 'machine - based system that can, for a given set of human - defined objectives, make predictions, recommendations or decisions influencing real or virtual environments'. • AI located and accessed in several ways: ( i ) centrally in data centres; (ii) centrally in the cloud; (iii) decentralised devices. Machine learning: Ability to learn Methods: Ability to reason Technologies: Physical enablement AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 9 Source: Deutsche Bank, OECD, OECD Legal Instruments . 4 TYPES OF AI based on memory; social and emotional intelligence; self - aware; big picture vs. task specific Reactive AI Limited memory Theory of mind Self - awareness Description No memory and are task specific, with no learning capabilities. They produce output based on the input they receive. Have memory capabilities, so they can use past experiences and previous input to inform future decisions and to make better predictions. The system would have the social intelligence to understand emotions. This type of AI would be able to infer human intentions and predict behaviour, a necessary skill for AI systems to become integral members of human teams. AI systems that possess a sense of self, which gives them consciousness. Machines with self - awareness understand their own current state in a way similar to how the human brain functions. Examples Deep Blue, the IBM chess program that beat Garry Kasparov in the 1990s. Deep Blue can identify pieces on the chessboard and make predictions but, because it has no memory, it cannot use past experiences to inform future ones. Some of the decision making functions in self driving cars are designed this way. Self-driving cars may in future have the capability to infer and react to changing emotions of their human occupants, as well as drivers and pedestrians around them. This type of AI does not yet exist, but there have been cases of semi sentient robots, for example, in 2020 at Columbia University a robot arm was able to build a model of its own body without prior knowledge. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 10 Source: Deutsche Bank, IBM, OECD.AI. Note: Skills are the 10 most prevalent AI skills among LinkedIn members worldwide, 2015 - 202 2, as identified by the OECD.AI. Similar/repetitive skills have been removed for succinctness. AI SKILLS AI Skill Description Use - cases Machine learning Where algorithms use historical data as input to predict new output values without being explicitly programmed to do so. Recommendation engines; business process automation Deep learning A subset of machine learning, where a neural network attempts to simulate human knowledge acquisition by processing large amounts of data. Automated driving; detecting cancer cells; automated hearing and speech translation Data structures Specialised format through which to store, organize, process and retrieve data in memory. For other types of AI innovation; implement computer programs Pandas An open - source Python package used for data analysis and machine learning tasks. For other types of AI innovation; computer programming Computer vision Extraction of meaningful information from digital images by computers trained to understand and interpret visual input. Automated driving; analysing x - rays; smart stores; road traffic analysis TensorFlow A free, open - source software library developed by Google for machine learning and AI. For other types of AI innovation; computer programming; Facebook's image recognition system; Apple's Siri Natural language processing Programming computers to understand and respond to text or voice data, including to writer's intent and sentiment. Request for information; translation; autocorrect; automatic text summarising ; chatbot Scikit - Learn A free, open - source machine learning library for Python. For other types of AI innovation; computer programming PyTorch A free, open - source machine learning framework based on the Torch library, part of the Linux Foundation umbrella. For other types of AI innovation; computer programming Neural networks Artificial neural networks mimic the human brain as programmed by a set of algorithms. Three neural networks makes a deep learning algorithm. Deep learning; marketing; search engine functionality AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 2. Today's AI landscape " " 11 "I believe there is no deep difference between what can be achieved by a biological brain and what can be achieved by a computer. It therefore follows that computers can, in theory, emulate human intelligence - and exceed it." - Prof. Stephen Hawking, theoretical physicist AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 12 CAPABILITIES: Today's top AI areas to watch Sources: Deutsche Bank, World Economic Forum, OECD.AI, Statista , Our World in Data, IBM. Machine learning - Machine learning involves building algorithms that learn from experience without being explicitly programmed to do so. - Publications in machine learning have more than doubled from 2015 to 2021 - Artificial intelligence / machine learning is the most popular speciality in US computer science PhDs, accounting for 21% of total graduates in 2020. Deep learning - A subset of machine learning, where a neural network attempts to simulate human knowledge acquisition. - Deep learning can make machine learning algorithms more efficient by eliminating some data pre processing as deep learning algorithms are able to process unstructured data - Examples include virtual assistants like Amazon's Alexa or Apple's Siri, chatbots, as well as medical purposes. At UCLA, deep learning helped researchers detect cancer cells in blood. Generative AI - Algorithms that generate new outputs like images, text, audio, based on the data they have been trained on and uses a type of deep learning called generative adversarial networks. - Generative AI has been a buzzword following the boom in popularity of ChatGPT and DALLE-2. Other examples include MidJourney and Codex. - Generative AI has seen significant adoption in US by the marketing and advertising industry (37%) and technology (35%). AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | CAPABILITIES: Machines are edging closer and closer to superhuman capabilities Source: De utsche Bank, Swedish Chess Computer Association. Note: The Elo rating system is a method for calculating the relative skill levels of players in games such as chess. - Machines have far surpassed even the greatest human chess champions, beating the highest Elo - rated human in 2006, and have since continued to improve to superhuman levels. - AI has not yet mastered complex language tasks, but the difference is closing fast, with humans performing only 1 percentage point better in abductive natural language inference in 2019. ELO rankings of best machine in chess (1984 - 2020) 13 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2006: Highest Elo - rated machine surpasses highest Elo - rated human. Machine Rybka 1.2 was rated 2,902, ahead of human Garry Kasparov at 2,851. 1984: Elo of a machine surpasses the average player, who has an ELO rank of 1,500. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | GROWTH: The number of global patent filings for AI technologies has grown over 50 - fold since 2010 Source: De utsche Bank, Our World in Data, WIPO . - Since AI emerged in the 1950s, over 340,000 AI - related invention applications have been filed. - But patents and inventions have accelerated in the last decade, rising from 2,560 patents in 2010 to over 140,000 in 2021. - The ratio of scientific papers to inventions has also decreased, from 8:1 to under 3:1 from 2016 onwards, indicating a new in ter est in creating commercial products rather than developing theoretical research. Annual number of patents filed for AI technologies 14 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Number of patent filings AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 15 GROWTH : China and India have seen significant advances in their artificial intelligence capabilities Contributions to all public AI projects Source: Deutsche Bank, OECD.AI , Top500. *Very high impact are contributions with over 100 forks. Note: this chart shows the share of contributions (i.e., "commits") made to AI projects (i.e. AI - related GitHub "repositories") by country and over time. AI project impact is given by the number of managed c opies (i.e., "forks") made of that project. • The US has historically led in AI research and software development, but has been losing ground as China and India have rapid ly advanced their research capabilities. • China has now closed the gap with the US in terms of contributions to very high impact public AI projects, producing 18% of a ll high impact contributions, vs 14% for the US. India now leads high impact contributions, producing 20% of the total. • China has improved its AI capabilities by become a world leader in AI publications, the quality of its AI research, its incre ase s in R&D spending in software and computing, and its development of supercomputers. China now possess 162 of the fastest supercomputer s, and the US has 127, as ranked by Top500. 0% 5% 10% 15% 20% 25% 30% 35% 40% 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 USA India EU27 China UK Japan Canada Other Contributions to very high impact public AI projects* 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 USA India EU27 China UK Japan Canada Other AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 16 GROWTH: And investment on the ground has been booming , driven by private investment and mergers and acquisitions Annual global corporate investment into artificial intelligence by category ($bn) Source: Deutsche Bank, Our World in Data. Note: Data for chart is expressed in constant 2021 US$. Inflation adjustment is based on th e U S CPI. • There is a global investment boom into AI, with total global corporate investment into AI up 150% since 2019, and nearly 30 - fold since 2013. 0 20 40 60 80 100 120 140 160 180 2013 2014 2015 2016 2017 2018 2019 2020 2021 $bns Private investment Mergers and acquisitions Minority stake Public offering Total corporate investment AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 17 GROWTH: Private investment has surged , focused on (i) data management, processing , cloud; and (ii) medical and healthcare Annual global private investment in artificial intelligence, by focus area ($bn) Source: Deutsche Bank, Our World in Data. Note: Data for chart is expressed in constant 2021 US$. Inflation adjustment is based on th e U S CPI. • This includes companies that received over $1.5m in investment. 0 2 4 6 8 10 12 14 2017 2018 2019 2020 2021 Billions Audiovisual Cybersecurity, data protection Data management, processing, cloud Financial tech Fitness and wellness Industrial automation, network Medical and healthcare Natural language, customer support Retail Semiconductor AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 18 LABOUR MARKET: Growth of the AI workforce was 63% from 2017 to 2019 in OECD countries, vs 3% for workers overall The share of total employment that possesses 'AI skills' Source: Deutsche Bank, OECD . Note: AI skills are those defined by Alekseeva (2021). Chart comes from OECD analysis of European Labour Force Survey (EU - LFS) for European countries, the Current Population Survey (CPS) for the US, and Lightcast data. • Across OECD countries, the average share of AI employment is just above 0.3%, ranging from 0.5% in the UK to 0.2% in Greece. • The AI workforce has grown rapidly from less than 0.1% in 2011 to above 0.3% in 2019 - due to increasing demand for AI skills wi thin occupations, rather than to the growth in occupations demanding AI skills. In other words, across almost all occupations, fir ms are demanding more workers with AI skills. • The occupations with the highest share of vacancies demanding AI skills are technical ones: mathematicians, actuaries and sta tis ticians (.52% demanding AI skill); software and application developers (4.9%); information and communication technology managers (4.3 %); database and network professionals (3.6%) and electrotechnology engineers (3.2%). 0.0% 0.1% 0.2% 0.3% 0.4% 0.5% UK Finland Sweden Netherlands Estonia Switzerland Israel Norway Denmark USA Belarus Austria OECD Germany France Lithuania Latvia Czech Republic Hungary Portugal Ireland Slovakia Spain Italy Greece 2011 2019 AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 3. What the opportunities are for corporates/banks 19 "It's very clear that AI is going to impact every industry. I think that every nation needs to make sure that AI is a part of their national strategy. Every country will be impacted." " - Jensen Huang, CEO of NVIDIA " AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 20 INNOVATION: Mentions of artificial intelligence in corporate documents surged by nearly 50% from 2020 to 2022 Source: Deutsche Bank, AlphaSense . Number of corporate documents mentioning "artificial intelligence" 10,000 15,000 20,000 25,000 30,000 35,000 Q1 2020 Q2 2020 Q3 2020 Q4 2020 Q1 2021 Q2 2021 Q3 2021 Q4 2021 Q1 2022 Q2 2022 Q3 2022 Q4 2022 Number of documents Company Doc Count Transcript Doc Count • AI has increasingly become a buzzword for corporates, with companies suggesting AI is being used in a range of sectors, from di stribution to medicine. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 21 BUILDING AI : Who? How much? Source: Deutsche Bank, IBM, The Financial Times, Bloomberg Finance LP, PR Newswire. Date Value Company Description March 2023 $22 million SK Networks South Korean firm SK Networks invested $22 million into Humane, an AI powered and software platform, in its Series C funding round. Humane looks to develop the next era of personal mobile computing powered by AI. March 2023 $25 million Salesforce Inc Salesforce Inc's VC capital arm is launching a $250 million fund in generative AI startups . Feb 2023 $300 million Google and Anthropic Google purchased AI startup Anthropic, a generative AI company. Anthropic has developed a rival chatbot called Claude, officially launched on March 14. Jan 2023 $10 billion Microsoft and OpenAI Microsoft has committed investing $10 billion into OpenAI over multiple years. This follows from its $1bn investment in 2019 and in 2021. 2022 $2 billion IBM IBM acquired eight companies for $2 billion in 2022, with 30 companies acquired in total since April 2020, to develop its hybrid cloud and artificial intelligence capabilities, including StepZen Inc in Feb 2023. It also invested $2 billion in an AI campus in 2019. • The release of ChatGPT last November and Microsoft's $10 billion investment into OpenAI in January have reinvigorated artificial intelligence investments. Recent investments have been made by Google, Salesforce and SK Networks. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 22 OPPORTUNITIES in artificial intelligence for corporates Sources: Deutsche Bank, Statista 31% 44% 40% 22% 34% 27% 46% 48% 31% 57% 14% 19% 20% 26% 26% 27% 27% 28% 29% 34% 0% 10% 20% 30% 40% 50% 60% Building brand awareness Increasing long-term customer engagement Increasing customer loyalty Reducing customer churn Acquiring new customers Recommender systems Detecting fraud Interacting with customers Retaining customers Improving customer experience % of respondents 2020 2021 AI and machine learning use - cases for companies worldwide • Use - cases for AI and machine learning are applicable for a broad portfolio of firms, typically used to create efficiencies in ex isting infrastructure as well as improving customer experience. • AI and machine learning have seen significant adoption for improving customer experience and interaction, as well as fraud de tec tion. • In the era of big data, the aggregation and analysis of large swathes of data will become increasingly processed by computers an d AI. • Even unstructured data can now be processed by deep learning algorithms, and transmute it into something digestible for other machine learning algorithms, as well as for human audiences. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 23 OPPORTUNITIES: For smaller corporates , the release of OpenAI's APIs will likely lead to a surge in innovation Cost for InstructGPT to process 120,000 words, as much as a typical adult fiction book contains* Source: Deutsche Bank, OpenAI , Bloomberg Finance LP. * InstructGPT is a model optimised to follow single - turn instructions. Note: Calculations are based on the assumption that one OpenAI token buys approximately 0.75 words. Experience Discovery Creator Economy • OpenAI's API was released in March including the API to ChatGPT - 3.5 and DALL - E. The API to ChatGPT - 4 is also now available. • For the 32k context model, it costs $0.06/1k tokens, and $0.12/1k tokens for completion, where 1,000 tokens is about 750 word s. • Adding AI capabilities to applications is now more accessible and affordable. Companies may emerge that focus on fine - tuning existing models rather than developing from the ground up. They are underpinned by technologies such as GitHub Copilot and Codex. • ChatGPT has now changed its data retention policy, only holding users' data for 30 days, and will not use input to train mode ls, reducing risks associated with trusting data to third - parties. • With the release of the official API, companies can charge for use of their products, and licensing is no longer a grey area. • Snapchat, Quizlet, Instacart and Shopify have all begun integrating the API. . 0 0.5 1 1.5 2 2.5 3 3.5 Ada Babbage Curie Davinci US$ Models ranked from fastest (Ada) to most powerful (Davinci) AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 24 Source: Deutsche Bank Research, OECD.AI, The AI Index Report 2022. RISKS: Technology is far from perfect - assistive but not a substitute • Accuracy: Risk both as creator and consumer of AI generated content - plausible text slipping into real submissions, unchecked and unverified (with no references), and maybe inappropriate for high - stake tasks. • AI decisions can be inscrutable and opaque; as they process more data, it becomes more challenging to document its details. • Meta launched Galactica, a science specific text generator, but was withdrawn three days later after producing fallacious and even dangerous information. • Bias: Larger language models are more capable of reflecting bias from their training data, and with greater capabilities, greater potential severity. • A 280 billion parameter model developed in 2021 shows a 29% increase in elicited toxicity over a 2018 117 million parameter model. • Cyclical industry with trend - chasers: • Global VC investment into AI has fallen as recessionary forces build, from over $200 billion to $120 billion. • Need to separate 'hype' vs meaningful value - add. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 3. Capabilities: Where we are today and where we are moving to 25 " " Machines take me by surprise with great frequency . - Alan Turing, founding father of modern artificial intelligence and cognitive science AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | AI related stocks have displayed strong PERFORMANCE since the beginning of 2023, far outperforming major indices Source: Deutsche Bank, Bloomberg Finance LP. Selected stocks based on Bloomberg Finance LP. *Updated 14 March 2023. Performance of selected AI exposed firms in 2023 YTD 26 • AI - exposed stocks have outperformed relative to the broader equity market, with players like C3.ai Inc. and NVIDIA Corp seeing r eturns up 89% and 65% respectively. • However, some companies have already started to falter as the risk rally ends. Buzzfeed grew 214% from the start to the end o f J anuary after announcing plans to utilise AI on its site, but the company has since lost nearly 60% of its value. 89 62 38 21 13 9 8 65 35 14 6 4 61 16 15 8 0 40 22 17 16 12 16 13 10 10 2 6 2 0 20 40 60 80 100 C3.ai Inc. Microstrategy Inc. Salfesforce Inc. Cloudflare Inc. Alteryx Inc. Microsoft Corp Soundhound AI Inc. Nvidia Corp Advanced Micro… NXP… QUALCOMM Inc Marvell Technology… 3D System Corp Metaage Corp Apple Inc Quadient SA Western Digital Corp Meta Platforms Inc SNAP Inc Baidu Inc Tencent Holdings LTD Netflix Inc Fair Isaac Amazon.com Inc Capgemini SE Tech Mahindra Ltd Cognizant Tech… STOXX 600 S&P 500 Software Services Media Hardware Semiconductors Major indices AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | FUTURE INNOVATION: Better, faster and more affordable? Source: Deutsche Bank, Our World in Data.*Cost in current USD of public cloud computing resources to train an artificial inte lli gence system to top 5 validation accuracy of 93% or more on ImageNet. ImageNet is a computer vision benchmark that has AI systems label images after interpreting visual input. 27 • Since 2018, the cost to train an image classification system has decreased by 63.6%, while training times have improved by 94 .4% . • The median price of robotic arms has decreased by 46.2% in the past five years from $42,000 per arm in 2017 to $22,600 in 202 1. • GPU computational performance per dollar has increased massively since 2006, from 62.42 million FLOP/s/$ with the GeForce 709 9 GX2 to 42.59 billion FLOP/s/$ with the GeForce RTX 3080 in 2020. • This will be favourable for corporates in their commercial adoption of AI. 1 10 100 1,000 2017 2018 2019 2020 2021 $, log scale Imagenet : Training cost (to 93% accuracy)* 0 10 20 30 40 50 60 70 80 90 100 2006 2008 2010 2012 2014 2016 2018 2020 Range from 0 to 100 (100=best) Protein folding prediction accuracy AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 28 INNOVATION : As computing demands grow , AI innovation is increasingly concentrated in industry , not academia Affiliation of research involved in the creation and building of 'notable' AI systems Source: Deutsche Bank, Our World in Data, Sevilla et al (2022). Note: 'notable' when the system is state of the art or of his tor ical importance. Experience Discovery Creator Economy • Intensity and computing demands increased with the prominence of deep learning, which has seen the size of ML systems and com put e demands increase, pricing out academic research. • Has occurred by shifting away from general purpose processors to more specialised hardware as ML systems predominantly train on specialised processors like Graphics processing units (GPUs), Tensor processing units (TPUs) and Neural Processing units (NPU s). • This may see AI increasingly opaque as backend of algorithms are produced within larger corporates in heated competition. 0 5 10 15 20 25 30 35 40 1950 1954 1956 1959 1961 1968 1974 1977 1980 1982 1984 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 Academia Collaboration Industry Collaboration, Academia-leaning Collaboration, Industry-leaning AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | FUTURE INNOVATION: Computational demands are increasingly significant, but talent will remain essential to future development Source: Deutsche Bank, OECD , OECD Digital Economy . 29 • On average, employment growth was 63% for the US AI workforce between 2017 and 2019, but just 3% for workers in the overall economy. AI employment had a 22% annual growth rate between 2011 to 2017. • The OECD argues that future advances in AI development will rely mostly on the human capital of its developers as well as on a d ynamic labour market for AI talent. • The average weekly hours of work for the AI workforce has grown by 0.7% over the sample period compared to a decline in avera ge weekly hours for the economy overall . • The computational capabilities required to train modern machine learning systems, measured in number of mathematical operati ons ( ie floating point operations per second, FLOPS) has multiplied by hundreds of thousands of times despite algorithmic and softwar e improvements that reduce computing power needs. Increasing demand for specialised AI software, hardware, related infrastructu re, as well as AI talent. Computational power required to train 'notable' AI systems ADALINE GoogLeNet / InceptionV1 LeNet - 5 MSRA (C, PReLU) Neocognitron Perceptron Mark I System 11 AlphaCode Minerva (540B) NPLM Fuzzy NN NetTalk 1E-14 1E-11 1E-08 0.00001 0.01 10 10000 10000000 1E+10 Jan 1941 Oct 1954 Jun 1968 Feb 1982 Oct 1995 Jul 2009 Mar 2023 PetaFLOP * Vision Language Games Speech Multimodal Drawing Other AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | SUSTAINABILITY: To achieve Net Zero, data centre emissions must halve by 2030, but are predicted to more than double Source: Deutsche Bank, OECD, OECD Digital Economy , International Energy Agency, Cambridge Centre for Alternative Finance. *This does not disaggregate the electricity use of a rti ficial intelligence from broader data centre electricity use. **Data for countries is for 2019, for global data centres is for 2020. ***Global data centre energy use excludes cryptocurrency mining. Annual electricity consumption for top 30 countries, and global data centres ( TWh )** 30 • AI is enabled by 'AI compute', the physical infrastructure, hardware and software required to run AI systems. • The IEA estimates data centre energy use has remained around 200 - 250 TWh , 1% of global electricity demand, despite growth in data traffic.* • However, growth in AI compute demands has outpaced the performance of current hardware, and it is unclear if this rate of efficiency gains can be maintained. • Andrae (2020) estimates data centre global electricity use may account for 783 TWh by 2030, or 2.6% of global electricity use. • And such a statistic neglects the environmental damage resulting from the mining of necessary rare materials, including that of rare earths, as well as transportation and end - of - life disposal. 0 2,000 4,000 6,000 8,000 China United States India Russian Federation Japan Brazil South Korea Canada Germany France Saudi Arabia Iran Mexico Italy United Kingdom Turkey Indonesia Global data centres Australia Spain Vietnam Thailand South Africa Egypt Poland Malaysia Ukraine Pakistan Norway Sweden United Arab Emirates Argentina TWh Production : raw material extraction; assembly; manufacturing Transport : distribution; freight transportation; handling and storage Operations : energy consumption; water consumption End - of - life : collection and shipping; dismantling and recycling; disposal Lifecycle of AI compute AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | SUSTAINABILITY: But AI can be used to improve sustainability Source: Deutsche Bank, OECD.AI, OECD, IEA. AI scientific publications related to the environment and energy use, by subject area 31 • AI can contribute to environmental action by improving efficiency and optimising the sustainability of existing systems, for exa mple, through climate mitigation and adaption technologies, environmental modelling and forecasting systems and technologies. • AI applications can also accelerate scientific research in the development of more sustainable technologies applicable to the gr een transition. • The number of AI scientific publications on 'Energy Engineering and Power Technology' and 'Renewable Energy, Sustainability a nd the Environment' have increased around ten - fold and forty - fold respectively. • The development of hyperscale data centres will also assist in reducing energy use, as they consume proportionally less energ y f or cooling relative to smaller data centres, according to the IEA. 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Number of publications Energy Engineering and Power Technology Renewable Energy, Sustainability and the Environment Earth and Planetary Sciences Atmospheric Science Environmental Science Nature and Landscape Conservation Waste Management and Disposal Water Science and Technology AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | SUSTAINABILITY: And data centres are now increasingly powered by renewable energy Source: Deutsche Bank, OECD.AI, OECD, IEA. Global renewable energy power purchase agreements by sector, 2010 - 2021 32 • ICT companies have increasingly invested in renewables where possible to reduce electricity price volatility and reduce carbo n e missions. • Amazon, Microsoft, Meta and Google are the four largest purchasers of corporate power purchase agreements, having contracted 15 gigawatts in 2021. 0 5 10 15 20 25 30 35 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Gigawatts ICT sector Other sectors AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | INNOVATIONS: Dividing the world between the haves and have - nots Source: Deutsche Bank, OECD , OECD Digital Economy , Top500 . Performance of supercomputers by country (Nov, 2022) 33 44% 13% 11% 7% 5% 4% 4% 2% 1% 1% 8% United States Japan China Finland Italy Germany France South Korea Russia United Kingdom Other 0 20 40 60 80 100 120 140 160 180 China United States Germany Japan France United Kingdom Canada South Korea Netherlands Brazil Italy Russia Saudi Arabia Sweden Australia Ireland Switzerland Singapore Norway India Finland Poland Taiwan Czechia Luxembourg United Arab Emirates Austria Slovenia Spain Morocco Bulgaria Thailand Belgium Hungary Number of supercomputers by country (Nov, 2022) • An imbalance of computer resources may deepen socioeconomic divides within and between countries, creating further difference s i n competitive advantage and productivity gains. • The Top500 list can serve as a proxy measure to observe emerging or deepening compute divides between economies - as supercomputers are increasingly updated to run AI specific workloads, gaps can be observed between those that have the resour ces to create complex AI models leading to productivity gains, and those that do not. • This is not merely about possessession , but about matching pace of development - nearly 90% of top supercomputers have been developed in the last 5 years, and their performance has grown 630 times in terms of computation capacity since 2009. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 34 REGULATION : Countries are increasingly introducing governance and regulation on AI, but continue to subsidise innovation AI policy instruments, 2021 Source: Deutsche Bank. OECD.AI , EUR - Lex, Our World In Data. *Countries selected are those with the highest number of AI - related bills as identified by Our Wor ld in Data. AI enablers and other incentives 27% Guidance and regulation 19% Governance 36% Financial support 18% AI enablers and other incentives Guidance and regulation Governance Financial support Cumulative number of AI - related bills (2016 - 2021)* • Examples of AI related policies: i. Scale - up policy - cutting - edge innovation in specific domains (e.g. health, transport and agriculture) ii. Scale - out policy: AI diffusion across sectors of the economy to unlock productivity gains and innovation at scale. • The EU classifies AI systems into three risk categories: i. Limited - risk: chatbots; gaming; inventory management ii. High - risk: pose risks to health and safety of persons; safety systems; consumer creditworthiness; employment decisions iii. Unacceptable - risk: cause societal harm; real - time, biometric identification used for public law enforcement; social scoring 0 2 4 6 8 10 12 14 2016 2017 2018 2019 2020 2021 Cumulative number of AI related bills US UK Russia China France Italy Japan South Korea AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 35 REGULATION: Countries have used a mixed portfolio of regulation to both enable AI innovation and to encode principles and ethics Source: Deutsche Bank, OECD , OECD.AI , EuroHPC , US Congress, The Federal Register. Countries Approach Goal European Union • Financial support. • Guidance and regulation. • Coordinated Plan on AI 2018 with a budget of more than EUR 500 million to maximise investments, cooperation. • The European High - Performance Computing Joint Undertaking ( EuroHPC ) was established in 2018 to share computing resources and coordinate efforts among EU countries and partners, with a 2021 to 2027 budget of EUR 7 billion. To build a European network of National Competence Centres for HPC. • Artificial Intelligence Act pending to promote beneficial development of AI and regulate autonomous weaponry. United Kingdom • Governance • Financial support. • 2021 National AI Strategy on access to finance for innovation, talent, international cooperation, public research capabilities through a National AI Research and Innovation Programme, Advanced Research and Invention Agency, increasing R&D to 2.4% of GDP by 2027. • October 2020, the UK announced launch of its most powerful supercomputer for use by healthcare researchers. United States • Governance • Financial support. • Guidance and regulation. • Executive Order: Maintaining American Leadership in AI 2019 focusing on tech developments, public and private collaboration, technical standards, talent, tech advantage, public trust. • Department of Energy launched the Frontier supercomputer as one of the world's most HPCs for AI applications in 2022. • National Science Foundation invests in next - gen AI R&D supercomputers such as Frontera (deployed in June 2019). • The US National AI Initiative Act of 2020 to make world - class computing resources and datasets available. • Algorithmic Accountability Act of 2022 to improve oversight of software, algorithms and automated systems. Japan • Guidance and regulation • Financial support. • Governance • In 2019, published the Social Principles of Human - Centric AI. • RIKEN Centre for Computational Science and Fujitsu launched top - ranked supercomputer, Fugaku , in 2020. • The National Institute of Advanced Industrial Science and Technology (AIST) develops and operates open AI computing infrastructure, including AI Bridging Cloud Infrastructure to accelerate AI R&D. China • Governance • Financial support • National New Generation AI Plan from 2017 seeks to make China 'world - leading' in some AI fields by 2025, to become the 'primary' centre for AI innovation by 2030 with AI industry worth RMB 1 trillion. • Partnership with national tech companies, and funding a USD 2.1 billion tech park in Beijing. India • Governance • Government stimulus • National Strategy on AI 2018 on proof - of - concept, AI ecosystem, collaboration with public, private and academic. • Centre of Excellence in Artificial Intelligence developing the National Artificial Intelligence Resource Portal. Offering to include a web - based system to search and browse AI resources, including training and cloud - based compute platform. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | 36 Latest research Thematic Research: US vs. China: Understanding the semiconductor chip war in 10 charts Thematic Research: US vs. China: Understanding the rare earth metals tensions in 10 charts Future Payments: CHARTBOOK - Stablecoins: Past, Present, and Future Future Payments: Asset Tokenisation, Blockchain & the Five Ws: Why, What, Who, When, Where? Future Payments: Everything you should know about the (coming) semiconductor chip war Future Payments: CHARTBOOK - The Future of Money AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | Appendix 1 Important Disclosures *Other information available upon request Analyst Certification *Prices are current as of the end of the previous trading session unless otherwise indicated and are sourced from local exchanges via Reuters, Bloomberg and other vendors . Other information is sourced from Deutsche Bank, subject companies, and other sources . For disclosures pertaining to recommendations or estimates made on securities other than the primary subject of this research, please see the most recently published company report or visit our global disclosure look - up page on our website at https://research.db.com/Research/Disclosures/FICCDisclosures . Aside from within this report, important risk and conflict disclosures can also be found at https://research.db.com/Research/Topics/Equities?topicId=RB0002 . Investors are strongly encouraged to review this information before investing . The views expressed in this report accurately reflect the personal views of the undersigned lead analyst(s) . In addition, the undersigned lead analyst(s) has not and will not receive any compensation for providing a specific recommendation or view in this report . Marion Laboure 37 Attribution I wish to acknowledge the contributions made by Anthony Chaimowitz of dbDIG Primary Research for his invaluable contributions to this report. AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | Additional Information The information and opinions in this report were prepared by Deutsche Bank AG or one of its affiliates (collectively "Deutsche Bank") . Though the information herein is believed to be reliable and has been obtained from public sources believed to be reliable, Deutsche Bank makes no representation as to its accuracy or completeness . Hyperlinks to third - party websites in this report are provided for reader convenience only . Deutsche Bank neither endorses the content nor is responsible for the accuracy or security controls of those websites . If you use the services of Deutsche Bank in connection with a purchase or sale of a security that is discussed in this report, or is included or discussed in another communication (oral or written) from a Deutsche Bank analyst, Deutsche Bank may act as principal for its own account or as agent for another person . Deutsche Bank may consider this report in deciding to trade as principal . It may also engage in transactions, for its own account or with customers, in a manner inconsistent with the views taken in this research report . Others within Deutsche Bank, including strategists, sales staff and other analysts, may take views that are inconsistent with those taken in this research report . Deutsche Bank issues a variety of research products, including fundamental analysis, equity - linked analysis, quantitative analysis and trade ideas . Recommendations contained in one type of communication may differ from recommendations contained in others, whether as a result of differing time horizons, methodologies, perspectives or otherwise . Deutsche Bank and/or its affiliates may also be holding debt or equity securities of the issuers it writes on . Analysts are paid in part based on the profitability of Deutsche Bank AG and its affiliates, which includes investment banking, trading and principal trading revenues . Opinions, estimates and projections constitute the current judgment of the author as of the date of this report . They do not necessarily reflect the opinions of Deutsche Bank and are subject to change without notice . Deutsche Bank provides liquidity for buyers and sellers of securities issued by the companies it covers . Deutsche Bank research analysts sometimes have shorter - term trade ideas that may be inconsistent with Deutsche Bank's existing longer - term ratings . Some trade ideas for equities are listed as Catalyst Calls ideas on the Research Website (https : //research . db . com/Research/), and can be found on the general coverage list and also on the covered company's page . A Catalyst Call idea represents a high - conviction belief by an analyst that a stock will outperform or underperform the market and/or a specified sector over a time frame of no less than two weeks and no more than three months . In addition to Catalyst Calls ideas, analysts may occasionally discuss with our clients, and with Deutsche Bank salespersons and traders, trading strategies or ideas that reference catalysts or events that may have a near - term or medium - term impact on the market price of the securities discussed in this report, which impact may be directionally counter to the analysts' current 12 - month view of total return or investment return as described herein . Deutsche Bank has no obligation to update, modify or amend this report or to otherwise notify a recipient thereof if an opinion, forecast or estimate changes or becomes inaccurate . Coverage and the frequency of changes in market conditions and in both general and company - specific economic prospects make it difficult to update research at defined intervals . Updates are at the sole discretion of the coverage analyst or of the Research Department Management, and the majority of reports are published at irregular intervals . This report is provided for informational purposes only and does not take into account the particular investment objectives, financial situations, or needs of individual clients . It is not an offer or a solicitation of an offer to buy or sell any financial instruments or to participate in any particular trading strategy . Target prices are inherently imprecise and a product of the analyst's judgment . The financial instruments discussed in this report may not be suitable for all investors, and investors must make their own informed investment decisions . Prices and availability of financial instruments are subject to change without notice, and investment transactions can lead to losses as a result of price fluctuations and other factors . If a financial instrument is denominated in a currency other than an investor's currency, a change in exchange rates may adversely affect the investment . Past performance is not necessarily indicative of future results . Performance calculations exclude transaction costs, unless otherwise indicated . Unless otherwise indicated, prices are current as of the end of the previous trading session and are sourced from local exchanges via Reuters, Bloomberg and other vendors . Data is also sourced from Deutsche Bank, subject companies, and other parties . The Deutsche Bank Research Department is independent of other business divisions of the Bank . Details regarding our organizational arrangements and information barriers we have to prevent and avoid conflicts of interest with respect to our research are available on our website (https : //research . db . com/Research/) under Disclaimer . Macroeconomic fluctuations often account for most of the risks associated with exposures to instruments that promise to pay fixed or variable interest rates . For an investor who is long fixed - rate instruments (thus receiving these cash flows), increases in interest rates naturally lift the discount factors applied to the expected cash flows and thus cause a loss . The longer the maturity of a certain cash flow and the higher the move in the discount factor, the higher will be the loss . Upside surprises in inflation, fiscal funding needs, and FX depreciation rates are among the most common adverse macroeconomic shocks to receivers . But counterparty exposure, issuer creditworthiness, client segmentation, regulation (including changes in assets holding limits for different types of investors), changes in tax policies, currency convertibility (which may constrain currency conversion, repatriation of profits and/or liquidation of positions), and settlement issues related to local clearing houses are also important risk factors . The sensitivity of fixed - income instruments to macroeconomic shocks may be mitigated by indexing the contracted cash flows to inflation, to FX depreciation, or to specified interest rates - these are common in emerging markets . The index fixings may - by construction - lag or mis - measure the actual move in the underlying variables they are intended to track . The choice of the proper fixing (or metric) is particularly important in swaps markets, where floating coupon rates (i . e . , coupons indexed to a typically short - dated interest rate reference index) are exchanged for fixed coupons . Funding in a currency that differs from the currency in which coupons are denominated carries FX risk . Options on swaps (swaptions) the risks typical to options in addition to the risks related to rates movements . 38 AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | Derivative transactions involve numerous risks including market, counterparty default and illiquidity risk . The appropriateness of these products for use by investors depends on the investors' own circumstances, including their tax position, their regulatory environment and the nature of their other assets and liabilities ; as such, investors should take expert legal and financial advice before entering into any transaction similar to or inspired by the contents of this publication . The risk of loss in futures trading and options, foreign or domestic, can be substantial . As a result of the high degree of leverage obtainable in futures and options trading, losses may be incurred that are greater than the amount of funds initially deposited - up to theoretically unlimited losses . Trading in options involves risk and is not suitable for all investors . Prior to buying or selling an option, investors must review the "Characteristics and Risks of Standardized Options", at http : //www . optionsclearing . com/about/publications/character - risks . jsp . If you are unable to access the website, please contact your Deutsche Bank representative for a copy of this important document . Participants in foreign exchange transactions may incur risks arising from several factors, including the following : ( i ) exchange rates can be volatile and are subject to large fluctuations ; (ii) the value of currencies may be affected by numerous market factors, including world and national economic, political and regulatory events, events in equity and debt markets and changes in interest rates ; and (iii) currencies may be subject to devaluation or government - imposed exchange controls, which could affect the value of the currency . Investors in securities such as ADRs, whose values are affected by the currency of an underlying security, effectively assume currency risk . Unless governing law provides otherwise, all transactions should be executed through the Deutsche Bank entity in the investor's home jurisdiction . Aside from within this report, important conflict disclosures can also be found at https : //research . db . com/Research/ on each company's research page . Investors are strongly encouraged to review this information before investing . Deutsche Bank (which includes Deutsche Bank AG, its branches and affiliated companies) is not acting as a financial adviser, consultant or fiduciary to you or any of your agents (collectively, "You" or "Your") with respect to any information provided in this report . Deutsche Bank does not provide investment, legal, tax or accounting advice, Deutsche Bank is not acting as your impartial adviser, and does not express any opinion or recommendation whatsoever as to any strategies, products or any other information presented in the materials . Information contained herein is being provided solely on the basis that the recipient will make an independent assessment of the merits of any investment decision, and it does not constitute a recommendation of, or express an opinion on, any product or service or any trading strategy . The information presented is general in nature and is not directed to retirement accounts or any specific person or account type, and is therefore provided to You on the express basis that it is not advice, and You may not rely upon it in making Your decision . The information we provide is being directed only to persons we believe to be financially sophisticated, who are capable of evaluating investment risks independently, both in general and with regard to particular transactions and investment strategies, and who understand that Deutsche Bank has financial interests in the offering of its products and services . If this is not the case, or if You are an IRA or other retail investor receiving this directly from us, we ask that you inform us immediately . In July 2018 , Deutsche Bank revised its rating system for short term ideas whereby the branding has been changed to Catalyst Calls ("CC") from SOLAR ideas ; the rating categories for Catalyst Calls originated in the Americas region have been made consistent with the categories used by Analysts globally ; and the effective time period for CCs has been reduced from a maximum of 180 days to 90 days . United States : Approved and/or distributed by Deutsche Bank Securities Incorporated, a member of FINRA, NFA and SIPC . Analysts located outside of the United States are employed by non - US affiliates that are not subject to FINRA regulations . European Economic Area (exc . United Kingdom) : Approved and/or distributed by Deutsche Bank AG, a joint stock corporation with limited liability incorporated in the Federal Republic of Germany with its principal office in Frankfurt am Main . Deutsche Bank AG is authorized under German Banking Law and is subject to supervision by the European Central Bank and by BaFin , Germany's Federal Financial Supervisory Authority . United Kingdom : Approved and/or distributed by Deutsche Bank AG acting through its London Branch at Winchester House, 1 Great Winchester Street, London EC 2 N 2 DB . Deutsche Bank AG in the United Kingdom is authorised by the Prudential Regulation Authority and is subject to limited regulation by the Prudential Regulation Authority and Financial Conduct Authority . Details about the extent of our authorisation and regulation are available on request . Hong Kong SAR : Distributed by Deutsche Bank AG, Hong Kong Branch except for any research content relating to futures contracts within the meaning of the Hong Kong Securities and Futures Ordinance Cap . 571 . Research reports on such futures contracts are not intended for access by persons who are located, incorporated, constituted or resident in Hong Kong . The author(s) of a research report may not be licensed to carry on regulated activities in Hong Kong and, if not licensed, do not hold themselves out as being able to do so . The provisions set out above in the 'Additional Information' section shall apply to the fullest extent permissible by local laws and regulations, including without limitation the Code of Conduct for Persons Licensed or Registered with the Securities and Futures Commission . This report is intended for distribution only to 'professional investors' as defined in Part 1 of Schedule of the SFO . This document must not be acted or relied on by persons who are not professional investors . Any investment or investment activity to which this document relates is only available to professional investors and will be engaged only with professional investors . 39 AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | India : Prepared by Deutsche Equities India Private Limited (DEIPL) having CIN : U 65990 MH 2002 PTC 137431 and registered office at 14 th Floor, The Capital, C - 70 , G Block, Bandra Kurla Complex Mumbai (India) 400051 . Tel : + 91 22 7180 4444 . It is registered by the Securities and Exchange Board of India (SEBI) as a Stock broker bearing registration no .: INZ 000252437 ; Merchant Banker bearing SEBI Registration no .: INM 000010833 and Research Analyst bearing SEBI Registration no .: INH 000001741 . DEIPL may have received administrative warnings from the SEBI for breaches of Indian regulations . Deutsche Bank and/or its affiliate(s) may have debt holdings or positions in the subject company . With regard to information on associates, please refer to the "Shareholdings" section in the Annual Report at : https : //www . db . com/ir/en/annual - reports . htm . Japan : Approved and/or distributed by Deutsche Securities Inc . (DSI) . Registration number - Registered as a financial instruments dealer by the Head of the Kanto Local Finance Bureau ( Kinsho ) No . 117 . Member of associations : JSDA, Type II Financial Instruments Firms Association and The Financial Futures Association of Japan . Commissions and risks involved in stock transactions - for stock transactions, we charge stock commissions and consumption tax by multiplying the transaction amount by the commission rate agreed with each customer . Stock transactions can lead to losses as a result of share price fluctuations and other factors . Transactions in foreign stocks can lead to additional losses stemming from foreign exchange fluctuations . We may also charge commissions and fees for certain categories of investment advice, products and services . Recommended investment strategies, products and services carry the risk of losses to principal and other losses as a result of changes in market and/or economic trends, and/or fluctuations in market value . Before deciding on the purchase of financial products and/or services, customers should carefully read the relevant disclosures, prospectuses and other documentation . 'Moody's', 'Standard Poor's', and 'Fitch' mentioned in this report are not registered credit rating agencies in Japan unless Japan or 'Nippon' is specifically designated in the name of the entity . Reports on Japanese listed companies not written by analysts of DSI are written by Deutsche Bank Group's analysts with the coverage companies specified by DSI . Some of the foreign securities stated on this report are not disclosed according to the Financial Instruments and Exchange Law of Japan . Target prices set by Deutsche Bank's equity analysts are based on a 12 - month forecast period . Korea : Distributed by Deutsche Securities Korea Co . South Africa : Deutsche Bank AG Johannesburg is incorporated in the Federal Republic of Germany (Branch Register Number in South Africa : 1998 / 003298 / 10 ) . Singapore : This report is issued by Deutsche Bank AG, Singapore Branch (One Raffles Quay # 18 - 00 South Tower Singapore 048583 , 65 6423 8001 ), which may be contacted in respect of any matters arising from, or in connection with, this report . Where this report is issued or promulgated by Deutsche Bank in Singapore to a person who is not an accredited investor, expert investor or institutional investor (as defined in the applicable Singapore laws and regulations), they accept legal responsibility to such person for its contents . Taiwan : Information on securities/investments that trade in Taiwan is for your reference only . Readers should independently evaluate investment risks and are solely responsible for their investment decisions . Deutsche Bank research may not be distributed to the Taiwan public media or quoted or used by the Taiwan public media without written consent . Information on securities/instruments that do not trade in Taiwan is for informational purposes only and is not to be construed as a recommendation to trade in such securities/instruments . Deutsche Securities Asia Limited, Taipei Branch may not execute transactions for clients in these securities/instruments . Qatar : Deutsche Bank AG in the Qatar Financial Centre (registered no . 00032 ) is regulated by the Qatar Financial Centre Regulatory Authority . Deutsche Bank AG - QFC Branch may undertake only the financial services activities that fall within the scope of its existing QFCRA license . Its principal place of business in the QFC : Qatar Financial Centre, Tower, West Bay, Level 5 , PO Box 14928 , Doha, Qatar . This information has been distributed by Deutsche Bank AG . Related financial products or services are only available only to Business Customers, as defined by the Qatar Financial Centre Regulatory Authority . Russia : The information, interpretation and opinions submitted herein are not in the context of, and do not constitute, any appraisal or evaluation activity requiring a license in the Russian Federation . Kingdom of Saudi Arabia : Deutsche Securities Saudi Arabia LLC Company (registered no . 07073 - 37 ) is regulated by the Capital Market Authority . Deutsche Securities Saudi Arabia may undertake only the financial services activities that fall within the scope of its existing CMA license . Its principal place of business in Saudi Arabia : King Fahad Road, Al Olaya District, P . O . Box 301809 , Faisaliah Tower - 17 th Floor, 11372 Riyadh, Saudi Arabia . United Arab Emirates : Deutsche Bank AG in the Dubai International Financial Centre (registered no . 00045 ) is regulated by the Dubai Financial Services Authority . Deutsche Bank AG - DIFC Branch may only undertake the financial services activities that fall within the scope of its existing DFSA license . Principal place of business in the DIFC : Dubai International Financial Centre, The Gate Village, Building 5 , PO Box 504902 , Dubai, U . A . E . This information has been distributed by Deutsche Bank AG . Related financial products or services are available only to Professional Clients, as defined by the Dubai Financial Services Authority . 40 AI & the five Ws: Why, What, Who, When, Where? | March 2023 Deutsche Bank Research | Australia and New Zealand : This research is intended only for "wholesale clients" within the meaning of the Australian Corporations Act and New Zealand Financial Advisors Act, respectively . Please refer to Australian specific research disclosures and related information at https : //australia . db . com/australia/content/research - information . html . Where research refers to any particular financial product recipients of the research should consider any product disclosure statement, prospectus or other applicable disclosure document before making any decision about whether to acquire the product . In preparing this report, the primary analyst or an individual who assisted in the preparation of this report has likely been in contact with the company that is the subject of this research for confirmation/clarification of data, facts, statements, permission to use company - sourced material in the report, and/or site - visit attendance . Without prior approval from Research Management, analysts may not accept from current or potential Banking clients the costs of travel, accommodations, or other expenses incurred by analysts attending site visits, conferences, social events, and the like . Similarly, without prior approval from Research Management and Anti - Bribery and Corruption ("ABC") team, analysts may not accept perks or other items of value for their personal use from issuers they cover . Additional information relative to securities, other financial products or issuers discussed in this report is available upon request . This report may not be reproduced, distributed or published without Deutsche Bank's prior written consent . Backtested , hypothetical or simulated performance results have inherent limitations . Unlike an actual performance record based on trading actual client portfolios, simulated results are achieved by means of the retroactive application of a backtested model itself designed with the benefit of hindsight . Taking into account historical events the backtesting of performance also differs from actual account performance because an actual investment strategy may be adjusted any time, for any reason, including a response to material, economic or market factors . The backtested performance includes hypothetical results that do not reflect the reinvestment of dividends and other earnings or the deduction of advisory fees, brokerage or other commissions, and any other expenses that a client would have paid or actually paid . No representation is made that any trading strategy or account will or is likely to achieve profits or losses similar to those shown . Alternative modeling techniques or assumptions might produce significantly different results and prove to be more appropriate . Past hypothetical backtest results are neither an indicator nor guarantee of future returns . Actual results will vary, perhaps materially, from the analysis . The method for computing individual E,S,G and composite ESG scores set forth herein is a novel method developed by the Research department within Deutsche Bank AG, computed using a systematic approach without human intervention . Different data providers, market sectors and geographies approach ESG analysis and incorporate the findings in a variety of ways . As such, the ESG scores referred to herein may differ from equivalent ratings developed and implemented by other ESG data providers in the market and may also differ from equivalent ratings developed and implemented by other divisions within the Deutsche Bank Group . Such ESG scores also differ from other ratings and rankings that have historically been applied in research reports published by Deutsche Bank AG . Further, such ESG scores do not represent a formal or official view of Deutsche Bank AG . It should be noted that the decision to incorporate ESG factors into any investment strategy may inhibit the ability to participate in certain investment opportunities that otherwise would be consistent with your investment objective and other principal investment strategies . The returns on a portfolio consisting primarily of sustainable investments may be lower or higher than portfolios where ESG factors, exclusions, or other sustainability issues are not considered, and the investment opportunities available to such portfolios may differ . Companies may not necessarily meet high performance standards on all aspects of ESG or sustainable investing issues ; there is also no guarantee that any company will meet expectations in connection with corporate responsibility, sustainability, and/or impact performance . Copyright © 2022 Deutsche Bank AG 41
37.5.0