Also selfsupervised learn is immensely important in reducing the amount of required data for supervised learning. Fx language models are to a large extend based on learning to predict the next token/word in a sentence. To do that well the model must have a good understanding of how to construct a sentence, paragraph etc.
Thx very interesting. I practice ML i generic term that encompass all kind of classical predictive models and classification fx linear regression, decision trees.
This is great. From what I've read about AlphaFold, it seems like it was devised using at least a heavy dose of supervised learning. Is it fair to say that most of the applications of AI that we've seen thus far (i.e,. GPT3, AlphaFold, DeepMind's main work) is based on supervised machine learning?
Thanks a lot. I am also on the private equity side, and thinking about focusing more on the public market. A great pleasure to learn from you on your journey to sector expertise.
Thanks for putting this text together, I think it is a very good introduction to the AI topic. I read a couple of books on AI last year (Super intelligence, Life 3.0, How smart machines think - I guess you read them too) and I found the subject fascinating. Looking forward to your next pieces which will be more business oriented I assume.
Great article. Although AI/ML is mooted as the next best thing, but seems marvellously theoretical. I'm still struggling to establish where AI can help businesses in practice. I work in banking / financial services and interested in actual case studies as to how data held on existing clients to find ways of generating additional revenue streams
One part of the primer will be on current and future use cases. It's not theoretical at all, it is being used in countless applications today and the future use cases are significantly more exciting.
All of these social media companies use AI. Drug developers are using AI. Banks / FI use AI for loan processing decisions. Autonomous driving, all AI. I will detail them in a later piece.
In the context of Banking, one simple example could be identifying the exact point in time that a client will or is likely to purchase a mortgage, using data points around transaction and savings behaviour. Once the machine has established this understanding, it will issue a trigger that sends a potential client an interesting offer for a mortgage or even earlier, help with navigating the home acquisition process.
Also selfsupervised learn is immensely important in reducing the amount of required data for supervised learning. Fx language models are to a large extend based on learning to predict the next token/word in a sentence. To do that well the model must have a good understanding of how to construct a sentence, paragraph etc.
Thx very interesting. I practice ML i generic term that encompass all kind of classical predictive models and classification fx linear regression, decision trees.
Excellent and informative write up Christopher. Look forward to the following write ups on this topic. Well done.
This is great. From what I've read about AlphaFold, it seems like it was devised using at least a heavy dose of supervised learning. Is it fair to say that most of the applications of AI that we've seen thus far (i.e,. GPT3, AlphaFold, DeepMind's main work) is based on supervised machine learning?
Thanks a lot. I am also on the private equity side, and thinking about focusing more on the public market. A great pleasure to learn from you on your journey to sector expertise.
Thanks for putting this text together, I think it is a very good introduction to the AI topic. I read a couple of books on AI last year (Super intelligence, Life 3.0, How smart machines think - I guess you read them too) and I found the subject fascinating. Looking forward to your next pieces which will be more business oriented I assume.
Very good job! Simple but still very informative. I am reading your articles and see how you develop yourself.
Congrats!
Great article. Although AI/ML is mooted as the next best thing, but seems marvellously theoretical. I'm still struggling to establish where AI can help businesses in practice. I work in banking / financial services and interested in actual case studies as to how data held on existing clients to find ways of generating additional revenue streams
One part of the primer will be on current and future use cases. It's not theoretical at all, it is being used in countless applications today and the future use cases are significantly more exciting.
All of these social media companies use AI. Drug developers are using AI. Banks / FI use AI for loan processing decisions. Autonomous driving, all AI. I will detail them in a later piece.
In the context of Banking, one simple example could be identifying the exact point in time that a client will or is likely to purchase a mortgage, using data points around transaction and savings behaviour. Once the machine has established this understanding, it will issue a trigger that sends a potential client an interesting offer for a mortgage or even earlier, help with navigating the home acquisition process.