Algos are still very simplistic, there are at least two in the market I mainly daytrade and although it's had some impact, the bulk of profits are still there to be made. Basically they pick off *some* (not all) of the "free money" from dumb customer orders. But 80-90% of trades still require pattern recognition that cannot be gleaned from simplistic analysis of the order book alone. I also game both algos when I need to exit, I sometimes get a few ticks better exit price by bluffing them into higher bids or lower offers.
Still, algorithms are very good for certain things like consistency, attendance, speed etc. For simple patterns with fat edge, and for arbs, they are better than a normal trader in many circumstances. Hence why I am now working on some myself, and I am sure most good daytraders are as well. Maths/computer whiz kids are cheap and plentiful, and algos are scalable across multiple markets, so just move with the times and adapt. There will always be a place for top-notch pattern recognition and so far humans are still the best at that.
There will always be progress, innovation, and change in the markets. The losers will make whining posts complaining about them, and lose their income streams and edge. The winners will adapt, use the change to their advantage, and thrive and prosper. This has always been the case in the past and will continue to be in future. You can tell a lot about someone by how they respond to and view change.