Machine learning is one of the most important trends in tech right now. But like anything new, it naturally raises a number of important questions and concerns. Benedict Evan's most recent blog post provides a good explanation of what he refers to as the artificial intelligence bias. Here are a couple of excerpts that I found interesting.
What machine learning does:
With machine learning, we don’t use hand-written rules to recognise X or Y. Instead, we take a thousand examples of X and a thousand examples of Y, and we get the computer to build a model based on statistical analysis of those examples. Then we can give that model a new data point and it says, with a given degree of accuracy, whether it fits example set X or example set Y. Machine learning uses data to generate a model, rather than a human being writing the model. This produces startlingly good results, particularly for recognition or pattern-finding problems, and this is the reason why the whole tech industry is being remade around machine learning.
The rub:
However, there’s a catch. In the real world, your thousand (or hundred thousand, or million) examples of X and Y also contain A, B, J, L, O, R, and P. Those may not be evenly distributed, and they may be prominent enough that the system pays more attention to L and R than it does to X.
What AI isn't:
I often think that the term ‘artificial intelligence’ is deeply unhelpful in conversations like this. It creates the largely false impression that we have actually created, well, intelligence - that we are somehow on a path to HAL 9000 or Skynet - towards something that actually understands. We aren’t.
The conclusion:
Hence, it is completely false to say that ‘AI is maths, so it cannot be biased’. But it is equally false to say that ML is ‘inherently biased’. ML finds patterns in data - what patterns depends on the data, and the data is up to us, and what we do with it is up to us. Machine learning is much better at doing certain things than people, just as a dog is much better at finding drugs than people, but you wouldn’t convict someone on a dog’s evidence. And dogs are much more intelligent than any machine learning.
Photo by Ales Nesetril on Unsplash
Fred Wilson made an interesting remark in his recent post about the current "IPO bonanza" that is taking place in the tech space. He is, of course, talking about the recent IPO of Lyft, the recent S-1 filings from Pinterest and others, and the expected filings from Uber, Airbnb, and so on.
After listing the benefits of going public, he went on to say that this bonanza will surely also mean that it is going to become even more unaffordable in the Bay Area. Part of this is perhaps self-serving, since he operates a VC firm out of NYC. (Take your money and move to NYC.)
But the data suggests that there is truth to this.
When Twitter when public in 2013, it was estimated that it created some 1,600 millionaires. This is great for the local startup ecosystem as many of these beneficiaries could go on to found their own companies and create a whole new batch of jobs. The money gets recycled.
But what does it do to the local housing market -- especially a supply-constrained one like that of the Bay Area where it is difficult to build?
In 2018, Barney Hartman-Glaser, Mark Thibodeau, and Jiro Yoshida penned a paper called, Cash to Spend: IPO Wealth and House Prices. In it, they looked at the impact of IPOs on local home prices in California from 1993 through to 2017.
What they found, among other things, was a "positive and significant association between local house price changes and firms going public." The price increases were also found to be the greatest the closer you get to the headquarters of the firm that just went public.
If you'd like to download a copy of the paper, you can do that here.
Earlier this month it was announced that the on-demand electric scooter and bike startup, Lime, had closed a $310 million series D round. This values the 18-month old company at around $2.4 billion and brings its total raise to $867.1 million. For comparison, Bird -- its main competitor -- has raised around $400 million.
These numbers should tell you about the kind of growth that the "micromobility" startup is seeing. They are now in 15 countries and its riders have taken over 34 million trips. In the last 7 months alone, the company reports that it has seen a 5.5x increase in ridership. They are seen as an affordable last-mile solution. Supposedly 1/3 of its users report an income of less than $50,000 per year.
Lime entered the Canadian market last fall via Waterloo. They have yet to expand anywhere else, though I suspect we'll see them in Toronto this spring/summer. One of the barriers is that their scooters (with airless tires) aren't equipped to deal with snow, so they currently pack them up during the winter months.
This is in addition to the regulatory challenges they are facing in cities all around the world. But like Uber, I am sure there is a compromise to be had.
Machine learning is one of the most important trends in tech right now. But like anything new, it naturally raises a number of important questions and concerns. Benedict Evan's most recent blog post provides a good explanation of what he refers to as the artificial intelligence bias. Here are a couple of excerpts that I found interesting.
What machine learning does:
With machine learning, we don’t use hand-written rules to recognise X or Y. Instead, we take a thousand examples of X and a thousand examples of Y, and we get the computer to build a model based on statistical analysis of those examples. Then we can give that model a new data point and it says, with a given degree of accuracy, whether it fits example set X or example set Y. Machine learning uses data to generate a model, rather than a human being writing the model. This produces startlingly good results, particularly for recognition or pattern-finding problems, and this is the reason why the whole tech industry is being remade around machine learning.
The rub:
However, there’s a catch. In the real world, your thousand (or hundred thousand, or million) examples of X and Y also contain A, B, J, L, O, R, and P. Those may not be evenly distributed, and they may be prominent enough that the system pays more attention to L and R than it does to X.
What AI isn't:
I often think that the term ‘artificial intelligence’ is deeply unhelpful in conversations like this. It creates the largely false impression that we have actually created, well, intelligence - that we are somehow on a path to HAL 9000 or Skynet - towards something that actually understands. We aren’t.
The conclusion:
Hence, it is completely false to say that ‘AI is maths, so it cannot be biased’. But it is equally false to say that ML is ‘inherently biased’. ML finds patterns in data - what patterns depends on the data, and the data is up to us, and what we do with it is up to us. Machine learning is much better at doing certain things than people, just as a dog is much better at finding drugs than people, but you wouldn’t convict someone on a dog’s evidence. And dogs are much more intelligent than any machine learning.
Photo by Ales Nesetril on Unsplash
Fred Wilson made an interesting remark in his recent post about the current "IPO bonanza" that is taking place in the tech space. He is, of course, talking about the recent IPO of Lyft, the recent S-1 filings from Pinterest and others, and the expected filings from Uber, Airbnb, and so on.
After listing the benefits of going public, he went on to say that this bonanza will surely also mean that it is going to become even more unaffordable in the Bay Area. Part of this is perhaps self-serving, since he operates a VC firm out of NYC. (Take your money and move to NYC.)
But the data suggests that there is truth to this.
When Twitter when public in 2013, it was estimated that it created some 1,600 millionaires. This is great for the local startup ecosystem as many of these beneficiaries could go on to found their own companies and create a whole new batch of jobs. The money gets recycled.
But what does it do to the local housing market -- especially a supply-constrained one like that of the Bay Area where it is difficult to build?
In 2018, Barney Hartman-Glaser, Mark Thibodeau, and Jiro Yoshida penned a paper called, Cash to Spend: IPO Wealth and House Prices. In it, they looked at the impact of IPOs on local home prices in California from 1993 through to 2017.
What they found, among other things, was a "positive and significant association between local house price changes and firms going public." The price increases were also found to be the greatest the closer you get to the headquarters of the firm that just went public.
If you'd like to download a copy of the paper, you can do that here.
Earlier this month it was announced that the on-demand electric scooter and bike startup, Lime, had closed a $310 million series D round. This values the 18-month old company at around $2.4 billion and brings its total raise to $867.1 million. For comparison, Bird -- its main competitor -- has raised around $400 million.
These numbers should tell you about the kind of growth that the "micromobility" startup is seeing. They are now in 15 countries and its riders have taken over 34 million trips. In the last 7 months alone, the company reports that it has seen a 5.5x increase in ridership. They are seen as an affordable last-mile solution. Supposedly 1/3 of its users report an income of less than $50,000 per year.
Lime entered the Canadian market last fall via Waterloo. They have yet to expand anywhere else, though I suspect we'll see them in Toronto this spring/summer. One of the barriers is that their scooters (with airless tires) aren't equipped to deal with snow, so they currently pack them up during the winter months.
This is in addition to the regulatory challenges they are facing in cities all around the world. But like Uber, I am sure there is a compromise to be had.
Share Dialog
Share Dialog
Share Dialog
Share Dialog
Share Dialog
Share Dialog
