Confessions of a researchaholic

The first rule, which you will also hear from others, is to choose a place that can attract top graduate students whom you can work with, because that is the main attraction for being a professor in a research university.

After that, consider other factors, like funding and geography (and thus why I went to HKU as a professor).

Question from someone about to graduate with a PhD:
Do you think I can reach out with our paper draft and videos to profs for postdocs? It’s still under review, but I kinda need to send the paper to make my case, and can’t wait any longer.

This is always a yes and no question and we have to make our own judgement. On one hand, we don’t want to let potential reviewers and recruiters feel that we are compromising review anonymity, which can hurt our paper submissions and job applications. On the other hand, we do want to present ourselves in the best possible way.

I never have this problem personally, as during my job hunts so far I have shared only public domain information.
However, if I have to add confidential materials (e.g., under peer reviews or patent applications), I will share vague information, such as an alternative paper titles with high level descriptions and a disclaimer that the work is under evaluation, and let the recruiters decide whether they want to ask for more.

External visibility is public domain information for everyone to see, such as publications and patents.
Internal reputation is often proprietary and known only inside our institution, such as design or implementation for a product.

Our internal reputation is more important for the particular jobs (e.g., performance reviews and rewards), but that often contains confidential information which only a few colleagues can vouch for.

Both are important for our careers. For example, when we want to switch jobs, most people can only evaluate our external visibility and will require references to vet our internal reputation.
Thus, it is important to maintain a good balance of both.
I am lucky enough to enjoy doing both (e.g., publications and products).

Without continual practice, some skills, once mastered, will mostly remain with us, such as driving a car or riding a bicycle, while others tend to atrophy, especially during the early learning stages.

In my observations for computer science, research requires constant engagement to keep top performance.
I stopped research publications around 2001 to 2003 while focusing on GPU architecture, which reduced my research output and took me significant efforts later to get back to form (e.g., single-author a paper during weekends and evenings).
Research (at least in CS) can easily become rusty probably because it involves many delicate skills, such as literature, ideation, algorithm, code, demo, writing, and presentation.
Even if we can more easily get back to some of these (e.g., coding, which feels like riding a bicycle to me), missing any others could still throw us off balance (in particular ideation and familiarity with state of art, which are related).

Thus, if you are a graduating CS student and you want to keep research opportunity open for your future career, it is important to find a first job that you can continue your research in some way.
(Not necessarily academia, as I think I can have more time doing research in a top industry lab than as an assistance professor.)
Otherwise, you might never be able to come back.