I would like to know how people avoid unnecessary competition if they are going to give a talk or present a poster about their work. Of course, presenting our work might hopefully lead to new collaboration and connections, however, it might lead to competition with others if they find the idea behind the work good. Do you only present your published work? or do you present your work when you are sure it is going to published soon so even if someone else wants to do the same work cannot finish in time?
Firstly, I think you overestimate the risk of someone stealing your ideas at a conference talk and publishing them ahead of you. Really, at the point when you're presenting your ideas at a conference, you are so much closer to a successful publication than most people:
- you have studied a lot in a chosen narrow field and understand the background of your problem really well
- you have a nice idea which you understand better than others
- you have a working code or set of experiments or other solid evidence of success
- you have made at least some literature review
- you have at least a draft of your paper in mind (and some plots already on your slides)
It is really unlikely some freshman in the area will understand your methods and results and publish them ahead of you. Also, bear in mind that most conference talks only hint the idea anyway, as 95% of details are impossible to fit and explain during the conference. And of course, the fact that you present something at conference is remembered by your colleagues, who might then be involved in the revision process. So my strategy is - the more publicity the better (but I am not prepared to make my git repository public yet, so don't ask).
However, if you really want to protect your contribution - only present what is published. Many people still do this, and they're mostly as successful as others.