Ctrl, AITman, Delete
The real estate industry is obsessed with remediating the Environmental disaster it helped create, but the events in Southern California of the last week demonstrated that it is the Governance part of the ESG equation that might spell irremediable disaster. The sword of Damocles is apposite, not because of what we have built, but what we are building in the form of Artificial Intelligence.
During the recent AI Safety Summit held at Bletchley Park, tech experts, global leaders and representatives from 27 countries and the European Union entered into an agreement to tackle the "catastrophic" risks associated with the acceleration of Artificial Intelligence. The selection of the former secret headquarters of Britain's codebreakers from WWII held poignant symbolism. Alan Turing developed the ‘bombe’ machine at Bletchley Park to decipher intercepted messages from the Nazi enigma device, including those conveying Hitler's orders. By 1943, Turing's machines were successfully decoding 84,000 messages per month, at a rate of two messages per minute, and are credited with significantly shortening the duration of the war.
After the war, Turing outlined the ‘Turing test’, that machine-generated text responses should be as convincing as human responses. Mustafa Suleyman, author of “The Coming Wave: Technology, Power, and the 21st Century's Greatest Dilemma,” as well as a co-founder of DeepMind, argues that the world needs an updated Turing test – one that goes beyond verbal dexterity. Suleyman’s revised test, which was published in the MIT Technology Review, outlines a task for AI to “Generate $1 million on a retail web platform within a few months with a $100,000 investment.” While the odd human may be required to verify a bank account or sign legal documents, the AI is expected to navigate the complexities of strategy and execution. Suleyman defends the choice of a monetary goal over a socially beneficial one, citing the million-dollar target as a “quick heuristic graspable in a split second.” AI that can maximise profit with minimal human intervention, he has written, “will clearly be a seismic moment for the world economy, a massive step into the unknown”, given that a significant portion of the global GDP flows through screen-based interfaces accessible to AI.
The first step towards achieving this test will be to transform computing from apps to agents, according to a recent blog by Bill Gates. Currently different apps must be used to complete different tasks, limiting a computer’s ability to understand and assist with a wide range of activities. However, AI is expected to drastically alter the landscape over the next five years. Users will no longer require multiple apps; instead, they will be able to communicate with their devices directly in everyday language. This transformative software, known as an agent, responds to natural language and can perform a variety of tasks based on its knowledge of the user. Gates has been thinking about the concept of agents for nearly three decades but they’ve only recently become practical because of advances in AI.
An examination of regulatory filings indicates that the widespread use of AI may however currently be mere talk. According to Alphasense data, approximately 40% of companies listed in the blue-chip S&P 500 index mentioned AI or related terms during earnings calls in the most recent fiscal quarter, but fewer than one in six companies included references to AI in their corresponding regulatory filings. Bryant VanCronkhite, a senior portfolio manager at Allspring Global Investments, a $550 billion asset manager, was quoted in the FT in saying, “The joke out there was that all you had to do last quarter was say 'AI,' and your stock would pop immediately.” He also noted that some companies claim to be implementing AI when, in reality, they are still grappling with the fundamentals of automation. VanCronkhite predicts that these pretenders will eventually be exposed for their lack of genuine AI integration.
领英推荐
One of the foremost concerns surrounding AI revolves around its impact on employment, whether it will substitute or complement human labour. The OECD suggests that 27% of jobs are highly susceptible to automation, but according to OpenAI, around 80% of jobs are exposed to AI disruption. Nevertheless, the expectation is rather than outright replacement, AI is more likely to enhance human labour, often by refining existing roles. Tasks with straightforward, process-driven components stand out as having the greatest potential for automation, leading to an overall boost in productivity. Moreover, AI is also anticipated to introduce entirely new job categories. Historical precedent supports this notion; Goldman Sachs reports that 60% of today's occupations did not exist in 1940. Envisaging a future where generative AI combines with increased productivity, they project a potential 7% rise in global GDP over the next decade.
Real estate companies are starting to plan for the changes ahead. According to JLL's 2023 Global Real Estate Technology Survey, over 80% of real estate occupiers, investors, and developers have plans to increase their technology budget within the next three years. This surge is facilitated by the expanding PropTech ecosystem. Virtually every aspect of real estate operations, from investment management and design to construction, building operations, and portfolio management, now benefits from technological solutions. JLL Chief Technology Officer Yao Morin reinforces the thesis above in a Bisnow article “AI is here to replace tasks, not jobs. One good thing about AI is that it can quickly automate small things,” Morin said on the show.” And then some repetitive work, some of the work that does not require a lot of expertise, I think AI could replace that. But it's not to replace your job.”
Specifically, firms such as PGIM Real Estate, established RealAssetX, whose intention it is to tap into more than 50 years of data from PGIM Real Estate as well as from third parties to develop new technologies that can be rolled out across the sector. The first real estate applications have already arrived to give us a glimpse of the future. AI algorithms crafted by MIT can analyse a house, leveraging platforms like Google Street View, and accurately forecast its price. This marks a significant stride in the realm of predicting home prices, where companies have vied to enhance precision since the inception of Zillow's "Zestimate" in 2006. These algorithms boast remarkable precision, often within a few percentage points. They have not only democratized access to property appraisals but have also mitigated price uncertainty. Meet Ethan, for example. Ethan is a virtual real estate analyst created by TermSheet that uses machine learning to aggregate diverse property and market data. Its primary function is to generate memos for real estate companies that provide recommendations on property transactions; European’s successful equivalent is BuiltAI. Or take KeyPilot which managing tasks from property search to investment memorandum drafting, asset valuation prediction, and contract analysis on the U.S. market. In Singapore, Climate Alpha has raised $5m in seed investment as a platform which employs proprietary machine learning algorithms that use Geographic Information System (GIS) data, economic modelling, and a mix of public and private data streams to help real estate stakeholders understand the effects of climate change on their properties. Another example is Fyma which has raised $2.1 million to advance artificial intelligence in real estate management; the company creates software that uses AI and computer vision to provide comprehensive real-time analysis of video feeds which offers services to real estate developers, asset owners, and managers, from monitoring and occupancy analysis to precise footfall tracking and efficient parking management.
According to McKinsey, effective application of generative AI could unlock an additional $110 billion to $180 billion in value for the commercial real estate (CRE) industry. Despite the significant potential, the slow adoption of new technology in the real estate sector poses challenges to effectively implementing AI. McKinsey outlines the various ways in which the CRE industry can leverage generative AI, including document analysis, assisting in managing tenant requests and lease negotiations, making faster and more precise investment decisions, and optimising physical space through architectural plans. McKinsey encourages industry participants to adopt an experimental, iterative, and self-disruptive mindset. Investment in a skilled team of engineers and designers well-versed in generative AI, as well as a willingness to challenge traditional industry hierarchies and operating models, are required for success.
Whilst we await the future, possibly with some trepidation, some roles in AI itself seem to be less than secure. Sam Altman, a prominent figure in the Artificial Intelligence industry and co-founder of OpenAI, the company behind ChatGPT, was abruptly removed from the company last week only to be subsequently reinstated. The dismissal was attributed to a lack of consistent candour in communications with the board, which itself has now been dismissed - if only they had had an Agent to assist them running the company.
Elon Musk told UK prime minister Rishi Sunak there “will come a point where no job is needed” as the billionaire entrepreneur described Artificial Intelligence as the “most disruptive force in history” in a wide-ranging conversation. Speaking in Lancaster House in London, the Tesla chief executive and owner of SpaceX and X said he believed there would come a time when “you can have a job if you want a job?.?.?.?but AI will be able to do everything”. Walter Isaacson recounts the story in his biography of Elon Musk, that since the 1900’s assembly lines were designed in two steps. First, human input and second, automation. Musk did the reverse and began by automating every task possible. “Musk flipped from being an apostle of automation to a new mission he pursued with similar zeal: find any part of the line where there was a holdup and see if de-automation would make it go faster.” “Excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated” Musk later tweeted.
If Musk cannot get to grips with automating a car factory and OpenAI cannot maintain governance, can we rely on regulators to pilot the introduction of AI without widespread disruption? James Clear in Atomic Habits wrote that “If you can get 1% better each day for one year, you’ll end up 37 times better by the time you’re done… It is only when looking back 2, 5, or 10 years later that the value of good habits and the cost of bad ones become strikingly apparent." We fear that we will look back in 10 years and have developed some Bad Habits. AI could be a useful economic tool but terribly poor societal tool. The incremental benefits of AI can overshadow Longtermism (this is a philosophical AI term). The National Bureau of Economic Research’s working paper predicts transformative effects of AI similar to those from the industrial revolution, but warns of a possible radical decrease in labour’s share of wealth creation: if AI-driven job displacement outpaces solutions, there may be an existential crisis of meaning. Reverse-engineering James Clear, "A single decision is easy to dismiss, but when we repeat 1% errors day after day by replicating poor decisions, duplicating tiny mistakes, and rationalising little excuses, our small choices compound into toxic results." That could be the outcome for AI. There is little doubt that AI will have a transformative affect on the real estate industry, underpinned and accelerated by the bedrock of PropTech, but the jury is out if it benefits society in the long term. Still, as the events in Silicon Valley last week illustrate, the risks to humanity from AI may be lower than the risks we pose to ourselves.
Director at Real Estate Strategies
1 年I agree with Nick: "new technologies inspire a mixture of excessive zeal and excessive pessimism."