Community
Since the launch of ChatGPT in late 2022, there has been a global surge in interest in AI, particularly in generative AI. Millions of dollars in venture capital are pouring into AI startups, and any tech company not incorporating AI is seen as outdated. AI has become essential not only for obtaining funding but also for securing customer interest.
While I am a huge fan of AI and use it almost daily, the current hype is clearly unhealthy. As I discussed in my blog "The Right Fit: Assessing Business Value before Adopting AI/ML" (https://bankloch.blogspot.com/2023/10/the-right-fit-assessing-business-value.html) the potential value and opportunities AI offers are enormous, but AI is not necessary or recommended for every use case. Assessing where AI brings value is essential, yet this crucial element seems to be often overlooked in today’s tech world.
This hype is reminiscent of the blockchain craze a few years ago. Banks were using blockchain for applications where it was unnecessary, but after the hype subsided, they now use it where it truly adds value. In my blog "Blockchain - Beyond the Hype" (https://bankloch.blogspot.com/2020/02/blockchain-beyond-hype.html") written during the peak of the blockchain hype, I predicted this evolution.
We can expect a similar trajectory for AI. I predict that the peak of venture capital investment in AIwill occur around the end of 2025, after which the market will correct itself, and AI will be used again where it genuinely brings value. Unlike blockchain, the AI hype is however significantly larger due to its broader range of potential use cases and greater potential value. Major tech companies like Google, Amazon, Meta, Apple and Microsoft/OpenAI have invested billions in AI, with plans to spend even more in the coming years (approximately $60 billion per year, according to Wall Street analysts).
The scale of venture capital investment in AI is just as staggering. AI startups captured a record 28% of global VC funding in Q2 2024, totalling $18.3 billion. Over the last five years, VC investments in AI have reached an estimated $290 billion. These figures demonstrate that the AI hype is on a much larger scale than the blockchain hype.
The stock market is also riding this wave. On July 31, 2024, NVIDIA, the leading provider of computing power for AI models, recorded the largest daily jump in market value in Wall Street history, adding $330 billion to its market capitalization. To put this in perspective, ASML, Europe’s largest tech firm by market cap, is valued at around $330 billion.
Big hypes ultimately lead to deep lows. Like the dot-com bubble, there will be significant impacts, but strong products and firms will probably still flourish. Analysts predict that when the market crashes, the market will be flooded with spare GPUs, impacting NVIDIA but also providing opportunities for the remaining players to acquire cheap hardware (for AI model training). Additionally, these remaining players will be able to acquire a lot of their competitors for a nickel and a dime, allowing to concentrate valuable expertise.
The dot-com bubble created many bubble companies but also boosted internet technology and led to the rise of major companies like Amazon, eBay, and Google. The AI bubble will be no different. With enormous investments and top tech talent, exponential advancements are happening, resulting in astonishing announcements almost daily. Hundreds of generative AI tools already exist, as I discussed in my blog "Beyond Imagination: The Rise and Evolution of Generative AI Tools." ("https://bankloch.blogspot.com/2023/11/beyond-imagination-rise-and-evolution.html").
In the meantime, the lists presented in this blog (less than a year ago) is already outdated, e.g. AI being used for
Research and agents: chatGPT, Anthropic Claude AI, Perpexity, Anakin.ai, websim.ai, Microsoft Copilot, Llama 3, Google Gemini, Mistral’s Le Chat, Bing Chat, Clearscope
Images: Lexica, DALL-E, Clarif AI, Segmind, Gencraft, Midjourney, Stable Diffusion, Leonardo AI, Playground AI, Jasper Art, NightCafe, AutoDraw, Designs.ai, Leap, Zapier,
Copywriting: Rytr, Crayon, Copy.ai, Surferseo, Wordtune, Writesonic, Jasper, Frase, Smart Copy by Unbounce, Writer.com, Anyword, Hypotenuse.AI
SEO: VidIQ, Alli AI, SerpStat, WordLift, BlogSEO, Seona AI
Chatbot: Droxy, Chatbot, Chatfuel, Chatbase, DialogFlow, ChatSimple
Presentation: Lumens, Gamma, Slides AI, Designs AI, Decktopus AI, Beautiful.ai, PopAI
Logo: Looka, Logo AI, Logaster, BrandMark, Namecheap, Stockimg AI
Audio: Udio, Lovo AI, LyreBird, Descript, Auphonic, ElevenLabs, Sonic,
Marketing: Adcopy, TryPencil, Sendbird, Simplified, AdCreative, MailChimp
…
At the same time astonishing accomplishment are published almost on a daily basis. Some recent developments in AI include:
OpenAI launches SearchGPT.
Meta releases Llama 3.1.
Mistral AI unveils Mistral Large 2 (ML2) and NeMo, a 12B model created in partnership with NVIDIA.
Chinese SenseTime unveils SenseNova 5.5, an enhanced version of its LLM.
Apple announces it will open-source its 7-billion parameter LLM model called DCLM.
Google’s experimental Gemini 1.5 Pro model surpasses OpenAI’s GPT-4o in generative AI benchmarks.
Google’s DeepMind AI wins silver at the International Math Olympiad and its AI weather model surpasses traditional forecasting.
New AI tools revolutionize medical diagnostics, including breast cancer (cfr. https://www.bcrf.org/blog/ai-breast-cancer-detection-screening/), dementia and Alzheimer’s ((https://www.nature.com/articles/d41586-024-02202-1) and prostate cancer detection (https://www.artificialintelligence-news.com/news/uk-hospitals-live-trial-prostate-cancer-detecting-ai/).
The financial services sector, being heavily regulated and risk-averse, is not aggressively adopting AI in core activities but is experimenting with its potential. Banks are increasingly partnering with Fintechs that heavily adopt AI. The biggest AI adoptions in financial services are seen in:
Security: AI-powered WAFs, risk-based authentication, and fraud detection engines
Compliance: Partnerships with RegTech firms offering AML, KYC, and sanction screening engines based on AI models
Risk Management: AI for predicting risks like credit scoring and improved liquidity risk management through better cash flow forecasting
Trading Algorithms and Market Research: AI models analyzing markets faster than humans, enabling quicker opportunity identification (e.g. JPMorgan’s AI chatbot, LLM Suite, which can do the work of a research analyst).
While adoption is considerable in these areas, other domains like chatbots, personalized financial recommendations (including robo-advice), and process automation are still in the experimentation phase. Clearly adoption is higher in back-end engines, where customers rarely interact directly with AI models. More sensitive areas involving direct customer interaction with the AI model are still in the experimental stage.
AI experiments are emerging in every sector, including financial services. These experiments often generate an initial "wow effect" but this excitement can quickly wear off as limitations, negative side effects, and bugs are identified. As a result, the aggressive customer-facing adoption of AI by banks and other financial institutions will take some time. However, given the current speed of innovation, this next phase of AI integration is on the horizon.
For more insights, visit my blog at https://bankloch.blogspot.com
This content is provided by an external author without editing by Finextra. It expresses the views and opinions of the author.
Ritesh Jain Founder at Infynit / Former COO HSBC
04 October
Nick Jones CEO at Zumo
Nkiru Uwaje Chief Operating Officer at MANSA
03 October
Dirk Emminger Managing Director at knowing finance
02 October
Welcome to Finextra. We use cookies to help us to deliver our services. You may change your preferences at our Cookie Centre.
Please read our Privacy Policy.