The Broad Dimension - 2023 Autumn/Winter

the newsletter of tbd consultants - Autumn/Winter 2023

Printable PDF version
Subscribe to our newsletter
In this Edition
Is AI After Your Job?
Rebuilding the Mall
Recession Receding?

Construction Management Specialists

San Francisco, CA
(415) 981-9430 (San Francisco office)

Orinda, CA
(415) 981-9430 (Orinda office)

Rocklin, CA
(415) 872-0996 (Sacramento office)

San Diego, CA
(858) 886-7373 (San Diego office)

Redmond, WA
(206) 571-0128 (Seattle office)

Los Altos, CA
(650) 386-1728 (South Bay office)

Los Angeles, CA
(424) 343-2652 (Los Angeles, CA office)

Wicklow, Ireland
+353 86-600-1352 (Europe office)

www.TBDconsultants.com

Is AI After Your Job?

AI has become the 70-year-old new-kid-on-the-block recently, as ChatGPT reignited interest in the subject. That led to the news media inundating us with warnings about AI being an existential threat that could wipe us out or that at least is going to take our jobs away. Much of that hype has been coming from industry leaders, although it has been suggested that their real motive for these warnings is to divert attention away from the fact that they are not revealing the source of the data that their systems are using. That data seems to include vast quantities of copyrighted material that the copyright holders are not being compensated for.

These AI systems can be a threat to jobs, and the IT industry is finding that out the hard way, because the recent language-model generative AI systems have proven to be very good at writing computer code. These systems analyze vast quantities of data to establish patterns, and then use those patterns to generate responses to queries. If the original data contains biases, those biases will also be reflected in the AI's responses. The biases could even become enhanced if the source data starts to include more AI-produced reports that already reflect such biases. The AIs are only looking at patterns in the data and are not understanding a single word or giving conscious thought to what is being said.

Could an AI become conscious? Nobody has a clue how we happen to be conscious, but our brains certainly have a big role in it. Brains are physical objects, just as computers are, so it's certainly feasible that a computerized AI system could become conscious someday. But that's not going to happen soon, even if it actually is possible. Computers can process data a lot quicker, and even more accurately than our brains can, but our brains are still vastly more complex than the most super supercomputer.

The above image was generated by AI

Being conscious lets us set goals for ourselves, while computers respond to input from users telling them what to do. If an AI was going to pose any sort of existential threat to us, it would be because some user set it the task of becoming such a threat. AI is a tool that we can use for good or evil, just like most types of tools.

As a tool, AI can increase productivity, and on the construction site that may just mean that contractors will once again have the means at their disposal to carry out the workload available to them. Currently they are hampered by a lack of trained workers, but with AI-driven robotic systems coming online and more advanced scheduling systems becoming available, that worker shortage will become less of a problem. AI systems can also monitor the site for safety issues and monitor and optimize the supply chain and inventory management. The construction industry is not known for its quick adoption of new technology, so the fact that such systems are already being brought into use on the site is an indication of the severity of the need.

How about at the start of the construction process, in the architect's design office? AI will certainly make it easy to produce optional concept designs that meet the client's needs, but what the AI will be doing is adapting previous designs that had been in its training data. It might also be combining aspects from one previous design with features from another and coming up with something that can count as original. However, true originality can only come from thought, and the AI has no conscious thought. Nevertheless, AIs will be able to produce multiple design concepts almost instantly and get the design process started. Once a concept is chosen, AI-based architectural systems will be able to work up design plans, look at options for space planning, structural and HVAC designs and the like, and it can generally speed the design process. It will also be able to review manually designed work and play a role in quality control, including checking the design against local building codes. This ability to review options can also lead to project cost reductions and give a more sustainable and efficient building for those using it.

That kind of help will, undoubtedly, reduce the need for architectural and engineering design staff, but it certainly won't eliminate them. There is a lot of "prototyping" in construction, and AI may be good at integrating existing data, but it's not so good when you need to extrapolate from the data to do something new. So, anything it comes up with will need to be reviewed and adapted as appropriate. In some cases, people are finding that it has been so easy for AIs to come up with options that they end up spending more time going through it all than they previously did while doing things manually.

Demonstrating the need for review, ChatGPT was consulted during the production of this article, and it seemed to imagine that architects spend a proportion of their time out on-site doing bricklaying. AI can frequently make more egregious errors than that, including causing a lawyer to quote fake court cases during a trial, so people will need to review the output. But editing can be more fun than doing the actual work. AI and humans will be watching each other, but the humans know where AI's off-switch is.

AI will undoubtedly change almost every field of construction, from design concepts through maintenance of the finished building, changing the way work is carried out and making it more efficient. That may actually help the industry in meeting the needs for additional new and alteration work imposed on us by climate change and business developments. During the life of the building, AI systems can also monitor energy usage, keeping conditions comfortable for users while minimizing the building's carbon footprint.

Rebuilding the Mall

Online shopping, followed by Covid, has had serious effects on shopping malls, and we take a look at how malls are adapting.

Recession Receding?

The financial markets are still in flux, and we examine how the odds are shifting for various outcomes; are we headed for recession, stagflation, or a soft landing?

Design consultant: Katie Levine of Vallance, Inc.



the newsletter of tbd consultants - Autumn/Winter 2023


		Printable PDF version Subscribe to our newsletter In this Edition Is AI After Your Job? Rebuilding the Mall Recession Receding?
	Construction Management Specialists San Francisco, CA (415) 981-9430 (San Francisco office) Orinda, CA (415) 981-9430 (Orinda office) Rocklin, CA (415) 872-0996 (Sacramento office) San Diego, CA (858) 886-7373 (San Diego office) Redmond, WA (206) 571-0128 (Seattle office) Los Altos, CA (650) 386-1728 (South Bay office) Los Angeles, CA (424) 343-2652 (Los Angeles, CA office) Wicklow, Ireland +353 86-600-1352 (Europe office) www.TBDconsultants.com


	Is AI After Your Job? AI has become the 70-year-old new-kid-on-the-block recently, as ChatGPT reignited interest in the subject. That led to the news media inundating us with warnings about AI being an existential threat that could wipe us out or that at least is going to take our jobs away. Much of that hype has been coming from industry leaders, although it has been suggested that their real motive for these warnings is to divert attention away from the fact that they are not revealing the source of the data that their systems are using. That data seems to include vast quantities of copyrighted material that the copyright holders are not being compensated for. These AI systems can be a threat to jobs, and the IT industry is finding that out the hard way, because the recent language-model generative AI systems have proven to be very good at writing computer code. These systems analyze vast quantities of data to establish patterns, and then use those patterns to generate responses to queries. If the original data contains biases, those biases will also be reflected in the AI's responses. The biases could even become enhanced if the source data starts to include more AI-produced reports that already reflect such biases. The AIs are only looking at patterns in the data and are not understanding a single word or giving conscious thought to what is being said. Could an AI become conscious? Nobody has a clue how we happen to be conscious, but our brains certainly have a big role in it. Brains are physical objects, just as computers are, so it's certainly feasible that a computerized AI system could become conscious someday. But that's not going to happen soon, even if it actually is possible. Computers can process data a lot quicker, and even more accurately than our brains can, but our brains are still vastly more complex than the most super supercomputer. The above image was generated by AI Being conscious lets us set goals for ourselves, while computers respond to input from users telling them what to do. If an AI was going to pose any sort of existential threat to us, it would be because some user set it the task of becoming such a threat. AI is a tool that we can use for good or evil, just like most types of tools. As a tool, AI can increase productivity, and on the construction site that may just mean that contractors will once again have the means at their disposal to carry out the workload available to them. Currently they are hampered by a lack of trained workers, but with AI-driven robotic systems coming online and more advanced scheduling systems becoming available, that worker shortage will become less of a problem. AI systems can also monitor the site for safety issues and monitor and optimize the supply chain and inventory management. The construction industry is not known for its quick adoption of new technology, so the fact that such systems are already being brought into use on the site is an indication of the severity of the need. How about at the start of the construction process, in the architect's design office? AI will certainly make it easy to produce optional concept designs that meet the client's needs, but what the AI will be doing is adapting previous designs that had been in its training data. It might also be combining aspects from one previous design with features from another and coming up with something that can count as original. However, true originality can only come from thought, and the AI has no conscious thought. Nevertheless, AIs will be able to produce multiple design concepts almost instantly and get the design process started. Once a concept is chosen, AI-based architectural systems will be able to work up design plans, look at options for space planning, structural and HVAC designs and the like, and it can generally speed the design process. It will also be able to review manually designed work and play a role in quality control, including checking the design against local building codes. This ability to review options can also lead to project cost reductions and give a more sustainable and efficient building for those using it. That kind of help will, undoubtedly, reduce the need for architectural and engineering design staff, but it certainly won't eliminate them. There is a lot of "prototyping" in construction, and AI may be good at integrating existing data, but it's not so good when you need to extrapolate from the data to do something new. So, anything it comes up with will need to be reviewed and adapted as appropriate. In some cases, people are finding that it has been so easy for AIs to come up with options that they end up spending more time going through it all than they previously did while doing things manually. Demonstrating the need for review, ChatGPT was consulted during the production of this article, and it seemed to imagine that architects spend a proportion of their time out on-site doing bricklaying. AI can frequently make more egregious errors than that, including causing a lawyer to quote fake court cases during a trial, so people will need to review the output. But editing can be more fun than doing the actual work. AI and humans will be watching each other, but the humans know where AI's off-switch is. AI will undoubtedly change almost every field of construction, from design concepts through maintenance of the finished building, changing the way work is carried out and making it more efficient. That may actually help the industry in meeting the needs for additional new and alteration work imposed on us by climate change and business developments. During the life of the building, AI systems can also monitor energy usage, keeping conditions comfortable for users while minimizing the building's carbon footprint.




	Rebuilding the Mall Online shopping, followed by Covid, has had serious effects on shopping malls, and we take a look at how malls are adapting.




	Recession Receding? The financial markets are still in flux, and we examine how the odds are shifting for various outcomes; are we headed for recession, stagflation, or a soft landing?




	Design consultant: Katie Levine of Vallance, Inc.