top of page
  • Writer's pictureeconomictribune

What can an 18th century painting tell us about Google’s AI bid?

By Konstantinos Paschalis, Economics Correspondent


In the halls of the National Gallery hangs a treasured painting called "An Experiment On a Bird in the Air Pump" by Joseph Wright, depicting a flurry of reactions to technological innovation. As the parrot is consumed by the contraption, one girl turns away in horror, whilst a man follows the process in a quiet awe.

"An Experiment On a Bird in the Air Pump" by Joseph Wright

Source: National Gallery


Two figures, silhouetted black against the widAlthough the painting was finished in 1768, the spectrum of opinions on technological breakthroughs today is not so different. A case in point is recent developments in generative Artificial Intelligence (AI). A select few such as Bill Gates, the founder of Microsoft, point towards a revolution in how humans access data, benefitting welfare and growth everywhere. Others anticipate mass joblessness or in the extreme case of Yuval Noah Harari, an anthropologist, a “hacking of the operating system of human civilisation.”


Clearly, the verdict is still out. What is certain, however, is that these advancements will lead to hefty gains for tech companies involved in the search engine market, a key application for AI algorithms. How might innovations rents earned from adopting AI alter the market's structure?


There is some historical precedence: In the late 90s when search engines were the new technological rage, Microsoft used its monopoly in operating systems, with over 90% market share, to make its own tool, Internet Explorer, dominant. As network effects and strong consumer habits were created by Windows' long running popularity, users surfed the web on Internet Explorer seemingly by default.


Today, Google holds an equally commanding 85% market share. But can it follow Microsoft's example by utilizing the same factors to ensure the success of its own AI, Bard? The answer is likely no, for two key reasons.


The first was put forward by Google itself. Earlier this year, a leaked employee memo titled "We Have No Moat" came to light. Having little to do with medieval castles, the memo instead described the extremely low barriers of entry to the AI market compared to other sectors. Although Google has developed Bard, if "one person, an evening and a beefy laptop" can yield a cutting-edge AI program, it leaves the company vulnerable to being left behind if it fails to adapt to a potentially more powerful algorithm.


Exacerbating the issue, Google's usual response of extensive investment may not be effective. Previously, if a new start-up was viewed as developing a service that was likely to rival any of Google's platforms in the future, they would just buy them out. In fact, 240 precautionary takeovers have been carried out since the beginning of the millennium.


If rival AI firms proliferate, this type of solution may be financially untenable. Admittedly, Google does possess huge scale advantages, meaning start-ups have a long way to go from their conception until they comprise a genuine threat in terms of market share. But, in the digital age, the popularity of substitutes can spread like computer viruses. Open AI's Chat GPT, which can in fact be seen as a substitute, reached over 100 million users within five days of launching.


The second reason is the possibility of government intervention. Internet Explorer was unprecedentedly dethroned through a landmark court case, as Microsoft was forbidden from tying it as the default search engine in Windows. This broke the cycle of inertia, pushing users towards preferring the most efficient alternative, google.com, which ultimately kick-started its road to dominance.


This event may form a precedent for a similar ban on Google, for example by making the promotion of Bard on its existing digital platforms illegal. For years Competition Commissions in the EU and USA have grumbled about Google's excessive market share. Thus, this specific ban, or something similar to it, would constitute a golden opportunity for regulators to throw a long-awaited sucker punch to curb Google’s monopoly powers to the AI newcomer’s benefit.


Therefore, due to low barriers to entry and government intervention, Google could be left in the dust. History repeats itself? Poetic, isn’t it?


The painting, "An Experiment On a Bird in the Air Pump" conveys another concept through the parrot's untimely demise. Innovation is often ugly and requires sacrifices.


Unknowingly, Wright has partially described what economists call creative destruction. When radical innovations come along, the old order, like the parrot, is sacrificed when firms who fail to adapt are forced to shut down due to profits drying up.


And Google is the AI race’s parrot. The optimal choice is flying away to safety if its scale advantages, like being large enough to own its AI development ecosystem, are significant enough to cement Bard as the superior software. Although the other outcome is not as grim as what happened to Wright’s parrot (not surviving), the firm will surely lose some feathers in the form of a decline in market share if Google’s executives are not vigilant to new competitors.




Flying away versus losing some feathers: The two outcomes for Google

Sources: Nature Conservancy Australia, Sydney Morning Herald


As Mark Twain allegedly said, “History does not repeat itself, but it often rhymes”. Will it this time? New search algorithms such as Perplexity, Andi and Komo are sprouting up around the web daily. Google’s response will be a deciding factor in their future. Sundar Pichai, Google’s CEO, can let competition fester or let Bard deliver a decisive blow.

 

References


1. “Bill Gates: AI is the most important technological advancement in decades” Gerken T., March 2023, BBC

3. “Facts About Microsoft”, Accessed 24/6/2023

4. “Worldwide Market Share of Search Engines” Bianchi T., May 2023, accessed 24/6/2023 https://www.statista.com/statistics/216573/worldwide-market-share-of-search-engines/

5. “We have no moat” SemiAnalysis, accessed 24/6/2023

6. “We have no moat” SemiAnalysis, accessed 24/6/2023

7. “Acquisitions by Google” Tracxn, accessed 24/6/2023

8. Chat GPT sets record for faster growing user base” Hu K., February 2023, Reuters, accessed 24/6/2023

9. “US vs Microsoft: Court Findings of Fact” The United States Department of Justice, accessed 24/6/2023

Comments


bottom of page