The Competitive Landscape for Machine Intelligence
1–10 of 13 posts
Re: The Competitive Landscape for Machine Intelligence
#2Regarding this - "Model here means business rules, like rules for approving loans or adjusting power consumption in data centers. In traditional software, programmers created these rules by hand. Today machine intelligence can use data and new algorithms to generate a model too complex for any human programmer to write."
Isn't it a bit problematic that the business rules generated by the model are too complex for humans to reason about them? How can you rely on the rules to be 100% appropriate for the task if it's impossible to reason about them?
Re: The Competitive Landscape for Machine Intelligence
#3Regarding this - "Model here means business rules, like rules for approving loans or adjusting power consumption in data centers. In traditional software, programmers created these rules by hand. Today machine intelligence can use data and new algorithms to generate a model too complex for any human programmer to write." Isn't it a bit problematic that the business rules generated by the model are too complex for hum…
Completely. In fact whenever regulators are involved (for example on anti money laundering), any black box model is a no go, precisely due to that inexplicability.
Re: The Competitive Landscape for Machine Intelligence
#4The PDF [0] is a bit of a grab bag and not as nicely organized as I would have expected from the past landscapes they've produced.
It's likely this coming from a less technical perspective but roboadvisors like Betterment or Wealthfront are not really examples of machine intelligence. Their whitepapers describe the techniques they use to craft their portfolios [1]. At best they use optimization on a predictive model, but it seems highly likely that they have manual input. They create a set of recommendations and execute them for you. There isn't much learning, data mining, or automated processing from data happening there.
The "Agent Enabler" section seems like its trying to get at foundational reinforcement learning companies but isn't self consistent with the examples provided.
They left companies like the Allen Institute and DeepMind off the research section.
It's easy to go on, but I think they need a technical editor next time :)
[0] https://hbr.org/resources/pdfs/hbr-articles/2016/11/the_stat... [1] https://research.wealthfront.com/whitepapers/portfolio-revie...
Re: The Competitive Landscape for Machine Intelligence
#5Regarding this - "Model here means business rules, like rules for approving loans or adjusting power consumption in data centers. In traditional software, programmers created these rules by hand. Today machine intelligence can use data and new algorithms to generate a model too complex for any human programmer to write." Isn't it a bit problematic that the business rules generated by the model are too complex for hum…
By "generate a model too complex for any human programmer to write" I believe the author is trying to say, to manually create the rules, one by one. Machine generated complex models, even though very complex, can definitely be understood and heavily audited.
One example is when you fit sparse high dimensional models to complex data in a real-time production system. The resulting models may have hundreds of millions to billions of features with non-zero weights, that constantly change as the underlying data changes. It's impossible to "hand-code" such a model from scratch by any reasonable size team in real-time. On the other hand, these hundreds of millions of rules can (and should) be exhaustively analyzed / audited by slicing and dicing both the model feature-weights, as well their performance on the data comprehensively. As an example, the "R" programming language typically creates useful human interpretable summaries for the models it generates.
For reference, I have been involved at Google in building such massive high dimensional models for properties like Youtube, and currently a founder of one the companies in the HBR report (LiftIgniter, YC W2014). Hopefully that doesn't make me too biased to respond.
Re: The Competitive Landscape for Machine Intelligence
#6The PDF [0] is a bit of a grab bag and not as nicely organized as I would have expected from the past landscapes they've produced. It's likely this coming from a less technical perspective but roboadvisors like Betterment or Wealthfront are not really examples of machine intelligence. Their whitepapers describe the techniques they use to craft their portfolios [1]. At best they use optimization on a predictive model,…
Go check out their past investments. It's a joke even by SV standards. It further highlights that they don't have anyone technical on hand that can point out obviously bad ideas. E.g. Gigster.
Gigster can't even filter out all of the bad customers that Godaddy intentionally sends their way. Every additional bad customer costs them in time, money and reputation. Yet they have the gall to claim they're AI powered. Amatures.
Re: The Competitive Landscape for Machine Intelligence
#7Regarding this - "Model here means business rules, like rules for approving loans or adjusting power consumption in data centers. In traditional software, programmers created these rules by hand. Today machine intelligence can use data and new algorithms to generate a model too complex for any human programmer to write." Isn't it a bit problematic that the business rules generated by the model are too complex for hum…
I think this is a major challenge too, although it depends. Some models are easy to reason about, e.g. Bayesian graphical models, while black-box approaches like deep neural networks are not.
One especially problematic issue is: if a model is too complex for humans to reason about, then a business could encode any kind of illicit behavior in the form of model parameters they like. Even if someone could prove that the model is biased one way or another, there is complete plausible deniability for the business, i.e. "I didn't make that choice, the learning algorithm did". We're in for some very interesting legal battles related to this, I think.
Re: The Competitive Landscape for Machine Intelligence
#8Under the "Audio" section, we've got Quirious, TalklQ, Twilio; melodious names that end with twirling, soft rhythm. Under "Internal Data", we've got Cycorp and Palantir and Primer; hard-edged P-prominent words implying secrecy and stead-fast solidity.
Re: The Competitive Landscape for Machine Intelligence
#9Regarding this - "Model here means business rules, like rules for approving loans or adjusting power consumption in data centers. In traditional software, programmers created these rules by hand. Today machine intelligence can use data and new algorithms to generate a model too complex for any human programmer to write." Isn't it a bit problematic that the business rules generated by the model are too complex for hum…
I think this is a major challenge too, although it depends. Some models are easy to reason about, e.g. Bayesian graphical models, while black-box approaches like deep neural networks are not. One especially problematic issue is: if a model is too complex for humans to reason about, then a business could encode any kind of illicit behavior in the form of model parameters they like. Even if someone could prove that the…
> if a model is too complex for humans to reason about, then a business could encode any kind of illicit behavior in the form of model parameters they like
If a model is too complex for humans to reason about, how would a business encode illicit behavior, unless the AI itself was running a significant portion of the business? Even in that case, someone or some group is responsible for setting the initial parameters of the model, and they can be held responsible for its decisions.
In the end, I think the legal solution would be put less emphasis on mens rea and more on actus reus. In other words, if your AI does something wrong, you are in the wrong, regardless of your intentions.
Re: The Competitive Landscape for Machine Intelligence
#10Regarding this - "Model here means business rules, like rules for approving loans or adjusting power consumption in data centers. In traditional software, programmers created these rules by hand. Today machine intelligence can use data and new algorithms to generate a model too complex for any human programmer to write." Isn't it a bit problematic that the business rules generated by the model are too complex for hum…
By "generate a model too complex for any human programmer to write" I believe the author is trying to say, to manually create the rules, one by one. Machine generated complex models, even though very complex, can definitely be understood and heavily audited. One example is when you fit sparse high dimensional models to complex data in a real-time production system. The resulting models may have hundreds of millions t…
If you're looking at model feature weights, you're doing it wrong.
Most models aren't interpretable, and coefficients are highly unit and feature dependent. Discussions involving feature weights beyond "What if we reduce the feature space?" or "Did we implement this feature correctly?" often go bad, they almost always go bad when you're using it to "audit" the model. I have been in way too many discussions where someone suggested that the weights were wrong, simply because they thought something should "be more important".