What you see here is the last week's worth of links and quips I have shared on LinkedIn, from Monday through Sunday.
For now I'll post the notes as they appeared on LinkedIn, including hashtags and sentence fragments. Over time I might expand on these thoughts as they land here on my blog.
Reminding us of the importance of the so-called "dull" parts of AI like … data pipelines and data supply chains:
"Google promised to delete sensitive data. It logged my abortion clinic visit." (WaPo)
I expect this is rooted in "incomplete data management practices" more than "nefarious intent."
But the end-result is the same: if you can't trace an element of data from source to destination(s), then you simply do not know where it is. You can't claim to control it.
So you certainly can't claim to delete it.
Dusting off my usual refrain of "AI is ideal for automating jobs that are all of Dull, Repetitive, and Predictable." Let's free up people for work that requires nuance and experience:
"Wendy’s, Google Train Next-Generation Order Taker: an AI Chatbot" (WSJ)
Hiring your company's first data scientist or machine learning engineer? You'll need the right mix of skills to build out the shop from scratch, bootstrap those data projects, and guide stakeholders/product owners on what to expect in an AI project.
Everyone in the data science/ML/AI space has their opinions on what to look for in that first hire. Here are mine: "experienced," "broad skill set," and a bonus of "some software development skills."
"What to look for in your company's first data scientist"
This post is part of a short series on hiring in data science/ML/AI. I ran the first one a few weeks ago: "Three questions to improve your data hiring"
For more posts on hiring, retention, and other such matters, check out the "employment" tag on my blog.
This article is a couple years old (which makes it practically ancient in online timescales) but everything in here still holds:
"Meeting the future: Dynamic risk management for uncertain times" (McKinsey)
This point, in particular, stands out:
Reset the aspiration for risk management
To meet the needs of the future, companies need to elevate risk management from mere prevention and mitigation to dynamic strategic enablement and value creation. This requires clear objectives, such as ensuring that efforts are focused on the risks that matter most, providing clarity about risk levels and risk appetite in a way that facilitates effective business decisions, and making sure that the organization is prepared to manage risks and adverse events.
Risk teams and stakeholders alike both need to see "risk management" as an enabler.
Risk management it's about saying "no" to everything. It's about spotting and addressing potential problems in advance, so you can steer clear of downside losses while leaving yourself open to upside gains.
A collaborative relationship with the risk management function can make this a reality.
When you're starting (or restarting) your company's AI efforts, it can be tough to know where to begin. Will AI even be useful for you? Is it even possible for your situation?
People ask me about this a lot. So I've pulled my thoughts together into a handy, one-pager website:
Work through just six questions to sort this out.
You can use this for yourself, you can pass it to your colleague who's very interested in AI, whatever … Have at it.
Just like my blog, "Will AI Help Here" isn't paywalled. It doesn't require registration. This is just a one-page, quick-reference website that I hope people will find useful.
(The only real catch is that you have to deal with my ancient HTML skills.)
I recently had the privilege of interviewing data scientist-turned-product-person Noelle Saldana for the Data Science Salon Podcast! We explored one of our favorite topics: companies that try to just "sprinkle a little AI" on a product.
Noelle shared her thoughts on why this happens, what to do about it, and how maybe -- just maybe -- there are ways to make it work.
The recent FAANG layoffs are not an automatic win for companies in search of tech talent:
"Not a Tech Firm? Snaring Top Laid-Off Tech Talent Won’t Be Easy" (WSJ)
https://www.wsj.com/articles/not-a-tech-firm-snaring-top-laid-off-tech-talent-wont-be-easy-7a6fdb2c
Key point:
In general, it is rare for a Big Tech worker to uproot for the IT department of a retailer, insurance company or other non-tech business, she said. “If Fortune 500 companies really want to compete for this talent, they have to ask themselves, how do we make this the kind of company where these people are going to want to work?”
This lesson isn't just for Fortune 500 companies. It's for every company.
I've noted before that hiring is a form of sales: your goal is to find someone who would likely be interested in what you have to offer, then show them why it'd be an improvement over their current situation.
What to look for in your company's first data scientist
Making the case for an experienced generalist with software skills.
Risk management for AI chatbots
Releasing an AI chatbot exposes your company to new risks. Here are some ideas on how to handle them.