We’ve all heard the scary stories, like the cyberattack in Atlanta (to name just one). We’ve all read the terrifying headlines: In Baltimore and Beyond, Stolen N.S.A. Tool Wreaks Havoc, I’m developing AI that can read emotions. It’s not as creepy as it sounds. (yes, it is…).

Image for post
Image for post

We’ve all seen the movies and TV shows where the AI wins (I’m talking about 100% of Season 2, Episode 3 of The Blacklist, 24 — basically all the seasons and at least half of all episodes, Seasons 2 through 4 of Person of Interest).

I think the question on everyone’s mind —…


The great statistician George Box worked in many areas — time-series analysis, Bayesian inference — and earned a handful of prestigious awards, on top of introducing the response surface methodology (RSM).

Image for post
Image for post
Statistician George Box, early skeptic of model usefulness

TL;DR: he was a really smart guy who knew his stuff. The kicker? He was probably best known for coining the expression “All models are wrong” back in the 1970s.

Now, you might be saying to yourself, “Yeah, but that was the 1970s. Now we have deep learning and AI and fancier technologies and stuff, and you can’t say that all those models on which our systems and lives…


Following OpenAI’s update from last weekend of the latest staged release of GPT-2, it’s a good time to reflect on what the project means for the future of AI’s openness, but more broadly — and probably more importantly — for the future of regulation and openness in the field.

Image for post
Image for post

Taking a step back for those that haven’t been following along, OpenAI is a nonprofit AI research organization co-founded by Elon Musk. They developed an AI called GPT-2, and with some impressive results. …


A Parallel History

Deep learning has met increasing hype in the last few years, and with lots of practical success. But does that necessarily indicate an exponential growth in AI over the next few years?

Amara’s law, stated by the futurologist Roy Amara, states:

“We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run”

In order to understand what’s to come, we need to figure out where we are exactly in the development of not only deep learning, but AI in general.

Image for post
Image for post
The Gartner Hype Cycle Curve

The Gartner hype curve gives some perspective about the adoption…


If you have kids, and if they are of the age where they have a phone, a tablet, a computer, or a video game console (or possibly all four of them), you may be — as a parent — scared. Like, very scared. Heck, even if you don’t have children, you might be scared by the idea of AI-dependent kids on the loose.

Image for post
Image for post

And, well, this fear might not be unfounded — screens could prevent children from creating social bonds or from fostering the kind of creativity that practically defines childhood. …


When it comes to the history of data teams and data-specific roles, we’re still very much in the wild west — that is, brand new territories where everything is kind of rough around the edges and (mostly) lawless. Sure, some best practices have come into focus in recent years, but for the most part, there isn’t one proven right way to do things; the fact that data professionals’ job titles (and the roles themselves) vary widely is further evidence of this.

Image for post
Image for post

One such fork in the road for the future of data teams, data roles, and even for AI as…


In most organizations today, having someone with a title like “Head of AI” (much less CAIO) would be a running joke. Mostly because, well… there aren’t a lot of companies out there yet who are really doing AI. But also because in theory, AI is something that should be distributed and pervasive across a business, not confined to a certain department or role.

Image for post
Image for post

Yet here we are: it’s 2019, and there are already people on LinkedIn with the title “Head of AI” (I won’t name names, but go ahead and search — you’ll see). …


Today, there (probably) isn’t a single company out there that would release a piece of software without testing it. Testing can ensure quality, save money, and eliminate risk. So it goes without saying that machine learning (ML) systems should also be tested.

Between 2016 and 2017, Google published two papers on building robust architecture for ML, including rubrics for self-assessment of one’s architecture. One of the underlying tenets of both papers (that, it’s worth noting, many companies spinning up their ML capabilities today tend to overlook or downplay) is the fundamental difference between testing software and testing machine learning systems.

Testing ML is different than testing software because of multiple steps (offline and online) and possible failure points (data needs to be tested, monitored, etc.). …


And How These Personalities Impact the AI Stack

“Chief architect” has long held meaning outside of the realm of the engineering and construction of physical buildings. Yet from managing different technologies or systems to interacting with and pleasing a range of different stakeholders, they still have many of the same challenges. And today’s chief architects are facing more of them than ever before in trying to build a high-quality system that will allow their company to seamlessly make the switch into the age of AI.

One of the most challenging aspects of building an AI stack is that, well —…


Reinforcement learning is a technique largely used for training gaming AI — like making a computer win at Go or finish Super Mario Bros levels super fast. That's great, but to some extent, this use case isn’t very exciting or useful. Yet reinforcement learning is also the potential first step toward bringing intuition to AI in business contexts, filling the gap between abductive reasoning and inductive reasoning.

Oh my! What does that last sentence even mean? Allow me to explain…

Image for post
Image for post

A Small Side Walk Though Epistemology

In philosophy, there are various ways to classify reasoning methods. But for the scope of this blog post, let’s focus…

Florian Douetteau

CEO @ Dataiku

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store