Most hiring processes start with an antiquated assumption that companies select people to work for them. In the world of technology startups we know that it’s usually the other way around: the best talent have plenty of optionality and the question is, how do you get *them* to choose you?

More recently though I’ve seen that signing an offer letter doesn’t always mean that person shows up on Day One. Counter offers and occasional ‘buyer’s regret’ have always been an issue, but a number of circumstances (including the shift to remote work) have anecdotally increased the renegs and no shows. As the 

 maybe some of it is a byproduct of tight labor markets,

“The rise in no-shows “could be just an expression of job seekers having a lot more confidence in their ability to find a job,” said Nick Bunker, an economist at the job-search platform Indeed.”

Or from the same article, perhaps it’s a change in cultural norms.

“We have a generation of professionals who grew up on dating apps, where ghosting has been accepted as an annoying, but common, phenomenon,” he [Keith Wolf, a managing director, Murray Resources] said. “I believe that is leaking into the professional world.”

You may be wondering if such a bold headline is true. The answer is yes. Let me explain why.

 came out in 2020 and established a new road the whole AI industry has been following in intention and attention since. Tech companies have repeatedly built better, larger models, one after another. But although they’ve put millions into the task, none of them has fundamentally changed the leading paradigm or the game’s rules GPT-3 laid out two years ago.

 (arguably the current podium of large language models) are significantly better than GPT-3 but they are, in essence, more of the same thing. Chinchilla has proved the success of slightly different scaling laws, but it’s still a large transformer-based model that uses a lot of data and compute, like the others.

, although distinct in what they do — text-to-image models that add techniques beyond the transformers — they’re pretty much based on the same trends. Even 

, which depart slightly from GPT-3 towards a more generalistic, multimodal approach to AI, are just a remix of the same ideas applied to novel tasks.

But, most importantly, all these AI models stem from the immense resources of private tech companies. That’s the common factor. It’s not just their technical specifications that make them belong to the same package. It’s because a handful of wealthy for-profit research labs exert absolute control over them.

The latest version of React Native now uses 

. Because React Native is somewhat coupled to React internals, React Native releases are generally tied to specific versions of React, and often official React Native releases lag behind the latest React version. With the release of version 0.69, React Native is now using the latest and greatest version of React.

Because of the tight coupling between the two, React Native has not always been able to take advantage of the most up-to-date versions of React, sometimes requiring developers to be a version or two behind. This means missing out on important performance and API improvements.

With React Native 0.69, developers can enjoy a number of new React 18 features like 

, a whole slew of bug fixes and performance improvements, and much more. This update is a critical step in making sure React Native is in sync with modern React development.

Many organizations start their data science journeys by overemphasizing AI 

 and gorging themselves on PhDs in their hiring frenzy. That’s a good idea if:

 and you’re not planning to ask your researchers what they’ve done for you lately.

You already know you’re solving a problem that is valuable, important, and 

known to be impossible to solve with all the existing algorithms out there

 is a high-risk investment and very few companies can afford it, because getting 

With the rise of no-code / low-code machine learning platforms and libraries, there are fewer barriers to use and apply machine learning models on your applications.

Citizen data scientists now have the ability to perform tasks that, in the past, would have needed a significant amount of coding experience thanks to low-code platforms and frameworks.

Although using a drag-and-drop interface to train a machine learning model on a no-code platform is the quickest and easiest method to do so, these platforms lack flexibility. On the other side, low-code ML is the optimal solution that strikes a balance between extremes.

In this article, I will discuss some of the most useful low-code machine learning libraries in Python. The list is subjective and is not organized in any specific fashion.

In comparison with the other open-source machine learning libraries, PyCaret is an alternative low-code library that can be used to replace hundreds of lines of code with only a few. This makes experiments exponentially fast and efficient. PyCaret is essentially a Python wrapper around several machine learning libraries and frameworks, such as scikit-learn, XGBoost, LightGBM, CatBoost, spaCy, Optuna, Hyperopt, and Ray.

All posts

Goodbye Excel. Hello Spreadsheets in Python!

More New Hires Are Ghosting Before Their Start Date. Here’s How To Prevent That.

BLOOM Is the Most Important AI Model of the Decade

What’s New in React Native 0.69 — How to Upgrade and Why it Matters

Why an AI researcher shouldn’t be your first data science hire

Low-code Python libraries for Data Science in 2022