The Futures - No. 05

From Quantumrun

May 16, 2023

TLDR: In this issue

The Quantumrun team shares actionable trend insights about China’s increasing electric vehicle (EV) export domination, the need for media streaming platforms to evolve, and the threat of open-source AI models to Big Tech.

China's electric cars are becoming Europe's biggest EV import market

Source: Car and Driver / Artist: Marcos Chin

According to Eurostat, 43 percent of Europe's 2021 electric car imports came from China, followed far behind by South Korea at only 20 percent. The growth of Chinese EV exports to the EU is not due to an improvement in the quality of the vehicles but rather a strategic shift by European and US automobile manufacturers to produce EVs in China for the European market. As barriers to market entry have decreased, these manufacturers have increased their regional investments.

Notably, Renault's Dacia Spring and the well-known Smart and Mini EVs from Daimler BMW are slated to be designed and manufactured in China for distribution worldwide. As of March 2022, Tesla remains the largest brand (49 percent) whose models are increasingly being produced in China, followed by Chinese-owned European brands (35 percent).

China's robust export growth is projected to persist as an increasing number of companies declare their plans for exporting and as Beijing gradually eliminates buying subsidies, a move that will likely temper local demand. In this context, Europe emerges as an especially appealing destination due to its presently minimal trade obstacles, well-established EV charging infrastructure, and generous EV purchase subsidies, which also apply to imported vehicles.

Image source: Carsalesbase, Eurostat, MERICS

China has long been a significant market for European automakers. They have historically established joint ventures in China to manufacture lower to mid-range vehicles, with most exports being luxury cars. The trade balance traditionally favored Europe, ranging between USD $3 billion and USD $5 billion. However, since 2020, there has been a shift. Europe still holds an advantage, but this surplus has dwindled to around USD $2 billion.

Actionable trend insights as Chinese EVs become the major European EV import

For job seekers: There are about 1.4 billion cars worldwide. Each year, roughly 80-100 million new vehicles are manufactured. This means it will take about 10-15 years to replace the existing global car fleet with EVs—and that assumes future generations will even need that many cars; they won’t (urbanization, AVs, declining driver license adoption rates, etc.) While there may be near-term job losses in the European automotive industry as European automakers struggle to compete with Chinese EV manufacturers, employment in this sector globally will remain robust throughout the 2020s. Moreover, battery production, factory robotics, vehicle software, Gigafactories, these are just some of the many innovations emerging around the auto sector that will create novel job opportunities well into the 2030s.
For entrepreneurs: Slapping a Western brand on low-cost, generic, Chinese-made products has been a common tactic for creating modest fortunes over the last two decades (e.g., dropshipping). This tactic may one day become possible with low-cost, generic, Chinese-made EVs. Western entrepreneurs could enter into white-label relationships with Chinese vehicle manufacturers as they scale their automated manufacturing facilities. As the cost of manufacturing EVs continues to plummet, new options for commodification will be inevitable.
For corporate innovators: The rise of Chinese EVs could shift the balance of power in the global automotive market. Rapid technological innovation in the EV sector is essential to stay competitive as Chinese manufacturers invest heavily in research and factory automation. This trend could lead to more advanced and affordable EVs being available to consumers worldwide.
For public sector innovators: European governments may face increased pressure to protect their domestic auto industries and jobs, leading to trade conflicts. For example, countries may impose carbon tariffs should China’s EV exports be increasingly produced using fossil fuels and other unsustainable practices.

Share The Futures

Future signals to watch

American businessman Mark Cuban thinks computer science—currently the world's highest-paying college major—will become less valuable to employers in the future.
"Reverse ATMs" are being installed by stores and restaurants as they transition to a cashless system that dispenses stored-value cards to customers in exchange for cash.
California has proposed a bill requiring all EVs, including school buses, to have bidirectional charging by 2027. This technology allows EVs to discharge power back into the grid during peak demand.
The US Food and Drug Administration (FDA) approved Arexvy, the world's first vaccine for respiratory syncytial virus (RSV). This scientific advancement has been 60 years in the making.
According to a Wall Street Journal-NORC April 2023 poll, 56 percent of Americans believe that the cost of a college degree outweighs its value (the highest percentage in the survey’s 10-year history).
A team of scientists has developed a language decoder that utilizes an artificial intelligence (AI) transformer similar to ChatGPT to translate a person's thoughts into text.
Enjay, a Sweden-based startup, has created a new type of heat exchange device named Lepido, which fills a gap in the market for capturing energy from kitchen exhaust fumes and redistributing it to heat other areas within a restaurant building.

Streaming wars need a new strategy to combat decreasing viewership

According to market research firm Nielsen, streaming remained dominant as of March 2023 when it came to viewership (34.1 percent). YouTube had the largest streaming viewership (7.8 percent), followed closely by Netflix (7.3 percent). However, a closer look at the trended views showed a more interesting picture. While YouTube views have steadily increased since July 2022, Netflix has gone the opposite way.

By February 2023, Nielsen reported that major streamers like Netflix, Hulu, and Amazon's share of US viewing time had decreased or plateaued since July 2022. YouTube (including YouTube TV) was the only platform that grew its viewing time during the same period.

Decreasing viewership can be attributed to subscriber churns (canceled subscriptions). According to former Amazon executive Roy Price, replacing subscriber churns are very costly for streamers. For instance, if a service has 50 million subscribers with an average revenue per user (ARPU) of USD $12 and 5 percent churn, it would cost them USD $200 million or 33 percent of their income to keep their subscriber count even.

For context, over 40 percent of Netflix's gross additions were "resubscribers," while Disney+, HBO Max, and Hulu each had about 30 percent. If Netflix had a 3 percent monthly churn, it would lose 23 million customers over 12 months, costing USD $1.8 billion to recover.

Actionable trend insights of streaming wars spurring new growth strategies

For job seekers: Even while budgets may be trimmed in the near term, the demand for innovative and affordable media properties will remain high, potentially creating job opportunities for young artists and media professionals.
For entrepreneurs: Startups that can create software and hardware solutions that reduce production costs for all types of media production will see heightened demand from media companies. Generative AI is one of many emerging technologies that can be adapted to simplify filming production processes.
For corporate innovators: With increased competition, streaming services may adopt a branded competition strategy, focusing on specific target audiences and consistently delivering content tailored to their interests. The target audience should be significant but not all-encompassing since customers come to know a service based on the type of content they provide. Since different people have different preferences, it's challenging to satisfy everyone. To stay competitive, streamers will have to choose target audiences and tones and cross-license content to other services when they own media that doesn't fit their brand.
For public sector innovators: Streaming platforms gaining more power and influence over what content is shown to viewers may spark growing concerns about censorship and bias. Governments may need to regulate streaming platforms further to ensure that they are not abusing their power and that content is not being unfairly censored.

Trending research reports from the world wide web

Leadership consulting firm DDI’s Diversity, Equity & Inclusion Report 2023 discussed how there had been a significant increase in businesses vowing to take concrete action towards Diversity, Equity, and Inclusion (DEI). Yet, only a handful have made considerable progress toward fulfilling their promises.
A recent survey conducted by the McKinsey Health Institute has discovered that while social media engagement can sometimes harm Generation Z, it also serves as a valuable tool for them to access mental health resources and foster connections.
A Georgetown University report discussed how the US government's shortcomings in swiftly securing software service contracts became evident in Ukraine. Not only did the Ukrainians quickly adopt the software, but it was apparent that digital services greatly influenced military activities.

Open-source AI models are about to outcompete Big Tech

A leaked document from a Google researcher discussed how vulnerable Google and OpenAI are to open-source AI models. The "major unsolved issues" of the past are now being effectively addressed and applied through these models. The researcher cited some examples, such as the Language Model on a smartphone that efficiently operates on Pixel 6 phones, processing five tokens per second.

Another example is the Scalable Individual AI, which allows fine-tuning a custom AI on a personal computer in just a single evening. There is also Ethical Distribution, where numerous websites now offer unrestricted access to artistic models, and the world of text-based models is not far behind. Finally, there are Multimodal Learning models like ScienceQA SOTA, which was trained in a mere hour.

While Google's models may still have a slight advantage in quality, the distance between Big Tech and open-source models is rapidly diminishing. Considering their size, these open-source models are quicker, offer more customization, ensure more privacy, and outperform Google. They're achieving results with USD $100 and 13 billion parameters that Google finds challenging with USD $10 million and 540 billion parameters. In addition, these breakthroughs are happening within weeks, not months.

The document reiterated that Big Tech has no exclusive knowledge or technique. Their best strategy is to learn from and engage with advancements outside their companies. They also need to give priority to third-party integrations. People won't be willing to pay for a restricted model when there are free, unrestricted alternatives of similar quality. In general, large-scale models are becoming a hindrance rather than an advantage. In the long term, the most effective models will be those that can be quickly iterated upon.

Actionable trend insights of open-source AI models outcompeting Big Tech

For job seekers: 2023 represents the hype phase of AI adoption. The next 10 years will involve the tedious but necessary work of implementing AI across corporations and factories to automate an infinite diversity of processes. This means that throughout the 2020s, there will be a growing demand for workers skilled in AI technologies and a corresponding proliferation of AI education and training programs.
For entrepreneurs: The growing adoption of open-source AI models will significantly lower barriers for startups entering the AI space, especially for startups applying AI to specific industries or processes. They no longer need to invest heavily in creating models from scratch, which can lead to the democratization of technology, foster competition, and lead to easier venture capital fundraising.
For corporate innovators: We may see Big Tech leaders buying out smaller AI startups to stifle competition and acquire new ideas. Nothing new here! That said, corporate tech leaders will continue to play a leading role in funding the cutting edge of AI innovation.
For public sector innovators: Policy advisors and regulators must remain vigilant as markets experience a potential worsening of existing biases if these LLM models are not carefully designed and implemented. This scenario could have significant implications in hiring, lending, and criminal justice, where AI models could unintentionally discriminate against certain groups.

Outside curiosities

The Alliance of Motion Picture and Television Producers (AMPTP) countered the Writers Guild of America (WGA)'s assertion that studios are trying to coerce writers into a "gig economy" and are planning to replace them with AI. The strike, which started on May 2, can significantly delay television series production and movie pipelines.
Sleep apnea and the inability to have deep sleep may increase the risk of dementia and stroke.
Wine and candy saved an Australian woman during a five-day ordeal in the wilderness.
A group of researchers at the New Mexico Institute of Mining and Technology in Socorro is transforming taxidermied birds into drones to investigate flight dynamics.