What fallacy is this?

technology experts have been saying technology will eliminate jobs for centuries now, and it just hasn't happened, So it never will."

Put more formally, it would look like this:

• Throughout history, there are those who have predicted that technology will become harmful to the workforce.
• Predictions of technology harming the workforce have constantly failed to come to pass since the beginning of technology.
• Therefore, such a prediction is unfounded and therefore false.

It might have something to do with the fact that past actions aren't an indicator of future actions that have yet to happen, but I'm still not quite sure what it would it even be called.

This appears to be a improper use of induction. Induction is great for making predictions but only probabilistically. For example, we can say that all past end of the world predictions have failed, so it is extremely likely (not with certainty—at least not based on the past examples) that a particular/current one will also fail (say 1,000,000 failed past predictions, and this is just one more, so 1/1million). However, to claim that all future predictions will fail (i.e., "never" come true) the mathematical equation has shifted from 1,000,000 past predictions in the last 5000+ years to one prediction, to 1,000,000 to perhaps a billion more in the next 100,000 years. This is unreasonable and an unwarranted conclusion.

In the context of the argument, the conclusion does not follow - non sequitur .

P1) People have been saying that X will happen.

P2) X hasn't happened yet (or has consistently failed to happen).

C) X will never happen (and concerns are unfounded).

This is a non sequitur. The fact people have made wrong predictions about X in the past does not mean X will "never" happen - circumstances may change to make X more likely, for instance, or even a certainty. So while we can be more skeptical of X, we can't simply rule it out.

This is even more fallacious if we claim X is unfounded, since we are wrong about both the truth value of X and its credibility as a claim. X may be very well supported, but just fail to materialise for one reason or another.

There's another implied (and I suggest false) premise that needs to be inserted into the "more formal" version of the argument.  That premise is that the accuracy of a current prediction is related directly to the accuracy of previous predictions.  It's related to the gambler’s fallacy .

It seems to me that the accuracy of a current prediction is more strongly based on the data and logical processes on which the current prediction is based ... not on the track record of previous predictions.  Assuming that predictions can be related because of their accuracy is a non sequitur in that the conclusion doesn't follow from the premises.

The best I get out of this argument is that basing future predictions on the same sort of data that were used for earlier predictions that didn't work out so well is likely to result in a new prediction that doesn't work out well, either.  It's instructive, but not in terms of whether to believe predictions; it's instructive (perhaps) in terms of the sort of data one needs to make good predictions.

Gambler fallacy. Something will happen or not based on probability.  We cannot say with 100% what will happen in the future only what has happen in the past.