The Futures - No. 04

From Quantumrun

May 09, 2023

TLDR: In this issue

We cover AI-based warfare strategy, how generative AI is making financial services more effective, the revamped LLM-fueled startups, and geopolitical foresight on the global recession.

Palantir launches AI-integrated platform for warfare strategy

Billionaire Peter Thiel’s tech company, Palantir, is launching its Artificial Intelligence Platform (AIP). This software is designed to run large language models, like GPT-4, on private networks. Palantir's pitch video showcases a military use of AIP, where an operator uses a chatbot to order drone reconnaissance, create attack plans, and organize the jamming of enemy communications. However, this vision raises concerns as it automates and abstracts military decision-making, with the human operator merely approving the chatbot's suggestions.

Palantir's AIP aims to provide a sense of safety and control for the Pentagon as it adopts AI technology. The platform claims to offer secure deployment on classified networks, control over AI systems, and a digital record of operations. However, AIP does not address the potential risks and consequences of using large language models in military situations, such as their tendency to generate false information. Instead, it focuses on providing "frameworks" and "guardrails" to ensure ethical and legal AI use in sensitive settings.

Palantir's CEO, Alex Karp, revealed that the company's data analytics software has played a significant role in Ukraine's targeting strategy against Russia since the invasion in 2022. The software helps Ukraine target assets like tanks and artillery. Palantir's technology allows for real-time tracking of the conflict and aids in decision-making based on enemy troop movements. Karp emphasized the importance of ethical considerations when deploying AI capable of independent action on the battlefield. In addition to its work in Ukraine, Palantir recently signed deals with Britain's Ministry of Defence and Japan's Sompo Holdings Inc.

Wider implications of Generative AI being integrated into warfare may include:

A potential for increased lethality, leading to unintended consequences and potential violations of the laws of war. Potential incidents may lead to a loss of public trust in military institutions, which could be worsened if AI systems cause unintended harm or are perceived as unethical.
A reduction in human casualties as autonomous systems could potentially replace human soldiers in certain roles, leading to fewer injuries and deaths.
The development and deployment of autonomous weapons systems becoming costly, potentially diverting resources away from other areas of national security and public welfare.
The integration of AI into warfare potentially affecting international relations and the balance of power between nations. There is a risk that countries with superior AI capabilities could dominate future conflicts.
Reduced number of soldiers required to carry out military operations, which could have implications for military recruitment and training.
Advances in AI and robotics, leading to new technologies and applications in civilian and military contexts.
More companies offering AI-integrated warfare strategy platforms to various defense organizations and paramilitary groups.

Share The Futures

Future signals to watch

Research showed that the workers with the lowest skills in customer service benefited the most when AI was introduced at Fortune 500 software companies for a year.
The US Food and Drug Administration approved the first tablet created from beneficial bacteria in human feces to combat harmful intestinal infections, offering a simpler method for fecal transplants.
The space around the Earth is getting more packed with satellites and debris, so space companies are brainstorming improved guidelines for managing outer space traffic.
France, the UK, Germany, Belgium, the Netherlands, Ireland, Denmark, Norway, and Luxembourg team up to accelerate North Sea wind farm expansion.
The interest in electric vehicles is surging, as sales are predicted to jump by 35 percent in 2023, following an outstanding performance in 2022.
PwC has partnered with Microsoft to sell Chat-GPT-based products on Azure to its IT consulting clients over three years. The deployment of this technology is expected to occur in the back-offices of large corporations like PwC, Accenture, and IBM over the course of the next decade.

Generative AI is becoming essential to financial services

Despite advancements in the field, fintech companies have had trouble making consumer finances easier to manage without human help, mainly because complex personal factors influence people's financial choices. This issue is evident when people decide which bills to pay when money is tight. Fintech tools struggle to provide the best advice in these situations. Similar problems are seen in managing investments and doing taxes, where human advisors do a better job since they can give more personalized advice. Large language models (LLMs) could help by understanding people's unique financial needs and making decisions based on them. This feature would change fintech from only helping customers in specific areas to offering tools to improve users' financial situations.

Meanwhile, the dream of using AI to make banking processes like mortgage approvals more efficient hasn't been realized yet due to three main challenges: consumer information being scattered across databases, financial services involving emotional decision-making, and strict regulations requiring human involvement. Generative AI can significantly improve these labor-intensive tasks by making data gathering, personalizing advice, and ensuring compliance more efficient. For example, customer service agents could access AI-trained models to quickly answer questions and reduce training time, while loan officers could use AI to instantly generate loan files from multiple systems. Additionally, AI can help speed up quality assurance and compliance checks, which can eventually result in instant access to mortgage approvals.

Generative AI has the potential to greatly improve the efficiency of compliance departments and help reduce the USD $800 billion to $2 trillion laundered worldwide annually. Compliance software is only 3 percent effective in stopping money laundering, and compliance employees spend much of their time dealing with false positives and gathering information. With generative AI, compliance officers could access critical information more efficiently. AI models could better predict money launderers by detecting new patterns in Suspicious Activity Reports (SARs). Large volumes of documents could be analyzed quickly to flag potential issues. Additionally, AI can be used to develop training materials and simulations for compliance officers.

Wider implications of generative AI in financial services may include:

Generative AI democratizing financial services by making personalized financial advice accessible to people from different socio-economic backgrounds. However, it could also exacerbate existing social divides if access to these advanced technologies is limited to a privileged few.
Generative AI improving efficiency and reducing costs in financial services, leading to increased profitability for financial institutions. This development could also contribute to a more dynamic economy by enabling better financial decision-making and resource allocation.
New regulatory challenges for governments. Policies are needed to ensure that AI is used ethically and transparently while balancing the need for innovation and economic growth.
Generative AI bridging the gap between generations in terms of financial literacy. By offering personalized financial advice, AI can help address the unique economic challenges different age groups face, such as retirement planning for the elderly or student loan management for younger people.
Further advancements in LLM technology, leading to a positive feedback loop, where improvements in AI capabilities further accelerate the adoption of AI.
Job displacement for some roles, such as financial analysts or customer service representatives. However, it could also create new job opportunities in AI development, maintenance, and oversight.
More sustainable investment strategies, as AI can analyze large amounts of data to identify environmentally friendly investment opportunities.
Ethical concerns about the potential for biases in AI algorithms, which could lead to unfair treatment of particular groups of people. Financial institutions and regulators will need to address these issues and ensure that AI-driven financial services are transparent, accountable, and fair.

Trending research reports from the world wide web

McKinsey analyzes the grocery sector in Europe amidst inflation and global economic uncertainty.
The revenue generated from PC and console sales amounted to USD $92.3 billion in 2022, representing a decrease of 2.2% when compared to the previous year. Although this decline may indicate a cooling down of the market following the pandemic, the industry's overall performance has exceeded our prior projections.
According to the Advertising Association / WARC Expenditure Report, the advertising market in the UK increased by 8.8 percent in 2022, even though there was a slight decline in the year's final quarter.

The startups of the future may be smaller but smarter

Language AI startups can be divided into several categories, and the first category is those that create and provide core general-purpose natural language processing (NLP) technology for other organizations to apply across industries and use cases (e.g., OpenAI). Building a state-of-the-art NLP model requires considerable resources and technical expertise, so very few companies develop their own NLP models. Most advanced NLP models today are based on a small handful of massive pre-trained language models, also known as "foundation models," built and open-sourced by publicly traded tech giants such as Google and Meta.

Another potential startup category is search services. Startups like You.com seek to disrupt the Google search experience by reconceptualizing the search engine from the ground up, emphasizing user data privacy. There are also significant opportunities for startups beyond consumer internet search, such as ZIR AI, which is building a new search platform for enterprises. Recent breakthroughs in AI have opened up opportunities for startups to build search tools for data modalities beyond text, particularly video. Twelve Labs is one such startup that uses cutting-edge NLP and computer vision to enable precise semantic search within videos, which will play a central role in AI's future.

Like lean startups, startups founded in the AI era may start on a small scale. They may leverage open source and cloud computing to begin their operations rapidly and iterate their products. However, due to the benefits of AI, these companies may remain small for extended periods, and the most successful among them will accomplish remarkable growth with only a few employees. Instances like Instagram, which had only 13 workers and was purchased for USD $1 billion, and WhatsApp, which had 35 engineers and supported 450 million users during its USD $16 billion acquisition, can become more prevalent. We will likely see firms with less than 100 employees going public.

Wider implications of LLM startup proliferation may include:

A shift in social interactions, as these models become integrated into daily communications, education, and media consumption. This trend could increase reliance on AI-generated content, potentially diminishing the importance of human creativity and the value of interpersonal skills.
Economic growth by creating new business opportunities and driving innovation in various industries. However, it may also lead to market consolidation as a few dominant players emerge and smaller competitors struggle to keep up with technological advancements.
Advancements in AI technology, leading to even more sophisticated and capable language models. This trend could result in a technological arms race as companies and governments compete to develop the most advanced AI systems for various applications.
The increasing demand for LLM services leading to greater energy consumption, as data centers and computational resources required for training and maintaining these models consume significant amounts of electricity. This development could contribute to increased greenhouse gas emissions and other environmental concerns.
Large language models can perpetuate existing biases and inequalities if not properly designed and monitored. Startups providing these services will need to address issues of fairness, accountability, and transparency to ensure that the benefits of AI technology are distributed equitably across society.

Outside curiosities

Apple emerged victorious in a court of appeals decision that supported its App Store's policies, as it faced an antitrust lawsuit initiated by Epic Games Inc.
Japanese tech investor SoftBank has initiated the sale of nearly all its remaining Alibaba shares, causing a decline in the Hong Kong-listed stocks of the Chinese e-commerce behemoth.
In Sweden, the Rehouse Niwa project, a collaboration between designers and the city government, uses a modular architecture and flexible city planning laws to fast-track construction. The 39-unit rental apartment building noted in the article is designed for easy disassembly and relocation after its 10 to 15-year temporary permit expires.
As the Arctic permafrost thaws and exposes buried mammoth skeletons, the ensuing demand for mammoth ivory risks endangering elephants once more.

Geopolitical foresight: The global economy is slowing down, as feared

The International Monetary Fund (IMF) recently released its least optimistic economic forecast in the last 50 years, predicting a global growth rate of only 2.8 percent in 2023 and 3 percent until 2028. One key reason behind this dismal outlook is the demographic decline in the developed world, where most economic activity is driven by private consumption. With lower birth rates and the retirement of the Baby Boomers, these countries will likely face reduced economic growth for the next 40 years.

Economic growth is also expected to stagnate in the developing world due to a lack of capital for building infrastructure and developing consumer markets. This capital traditionally came from developed nations, but with an aging population requiring more resources for pensions and healthcare, the funds are no longer available. Meanwhile, China, which has maintained high growth rates for decades, is facing demographic challenges and record debt levels, as well as losing its de facto American sponsorship for risk-free, cheap access to global markets.

Many emerging economies are also facing rising public and private debt levels, with middle-income countries at record highs and low-income countries nearing debilitating levels. Over 10 percent of the world's low-income countries face unsustainable debt burdens, and 50 percent risk joining them. Additionally, Russia's invasion of Ukraine has caused significant increases in food and energy costs, impacting importing countries' fiscal costs, draining reserves, and slowing growth. Rising prices pressure central banks in the rich world to tighten monetary policy.

As the globalized world unwinds geographically and demographically, lower growth rates will become the norm. However, this doesn't mean that every country will suffer equally. While some countries like Vietnam, Mexico, and the US may benefit from nearshoring and reshoring of manufacturing, others like China, Korea, and Germany may face significant economic challenges. This shift towards a "global starvation diet" in terms of economic growth may reshape the global landscape in the future.

Wider implications of addressing green energy economics may include:

Lower GDP growth rates, reduced foreign investment, and increased national debt levels. These trends may decrease government spending on public services, infrastructure, and social programs, further worsening economic inequality.
Political instability, as citizens become increasingly dissatisfied with their respective governments' inability to address economic issues. This development could result in the rise of populist movements, increased protectionism, and a shift in the balance of global power.
High unemployment rates and reduced economic opportunities encouraging individuals to migrate to other countries in search of better prospects, while low-income families choosing to have fewer children due to financial constraints.
Limited investment in research and development, slowing the pace of technological innovation. This trend may hamper the development of new technologies and industries, which are crucial for maintaining long-term economic growth and improving living standards.
Increased unemployment and underemployment as businesses struggle to maintain profitability and growth. This development can result in wage stagnation, reduced job security, and a shift towards more temporary and part-time work, negatively affecting workers' overall well-being.
Reduced industrial production and consumption might decrease pollution and resource depletion. However, governments and businesses may deprioritize environmental concerns in favor of short-term economic gains, reducing investment in sustainable practices and technologies.
Strained international relations and cooperation as countries become more focused on addressing domestic issues and protecting their interests. This move could hinder global efforts to tackle pressing challenges such as climate change, poverty reduction, and international security.