One key thing to remember is that history is in effect 'telescoped'- that is to say, things that are closer to us in the present seem bigger and more important than things further back. As an example, take the invention of metal. There is a general cultural perception, reinforced by popular media, that 'the stone age' was a fairly short, boring period of our history, where nothing much happened. Then we discovered how to produce metal and we were off to the races! Advancing into the period of 'real', 'meaningful' history.

Yet if you actually look at the timescales involved, early humans were using stone tools for millions of years. The portion of human history from the first discovery of metals to us sitting here using the internet is actually only a miniscule fraction of the whole. Yet to us, it contains most of the things we consider to be worth talking about. The main reason for this, is that we can only talk about things that we know about, and we can only know about things that have survived one of the most destructive forces on the planet - time.

We talk about things that are closer to us, because more information, knowledge and artefacts have reached us from those eras. Think about the huge place the second world war occupies in our historical discourse, despite only taking up a very short time, chronologically speaking. Now partly that's to do with it being such a monumental, global event, but it's also because there are still living people who remember it, so it forms a living part of our social discourse. Yet even it is currently fading as it passes from living memory into the realms of pure history.

Next in line we have the wider historical period, where humans had invented writing and began to document at least some aspects of their world (the ones they felt were important anyway) - even this period fades away the further back we go - written records degrade with time like all other physical objects (paper more so than most), and even where the earliest written languages survived (say on clay tablets), for some of them we have lost the means to read them. They too have been lost to history.

I give the above viewpoint to make two main points relevant to your question. One, we have a naturally biased viewpoint on the importance and duration on past eras, and two, that time itself destroys historical artefacts, records and knowledge, reinforcing the first problem.

To take your example of weapons, when we look at say, all the bronze swords that have been found. We cannot say 'this is an accurate image of weapons of this period'. The most we can say is that 'this is an example of some of the technologies of the time that have survived to us' and that is an entirely different statement. We can use it as an example of some of the things that an ancient society was doing, but we can never, with 100% confidence say that is all they were doing.

Take the example of the Antikythera mechanism, an amazing example of both metalwork and mechanical engineering found in an ancient Greek shipwreck. Before we found it, we had no conception that the Greeks of the period had such skills or knowledge. Why? Because no evidence of such technology had survived to us. Even after finding it, we have no idea how commonplace something like this was. Was this machine a lone work of genius, a mechanical marvel like no other? Or is it the lone survivor of an extremely common device that could have been found in any Greek town? We'll probably never know the answer - like most of humanity's existence, time has stolen it away. Metal's ability to survive for a great length of time is one of its great characteristics. Sadly, it's ability to be melted down into new objects means there are many ancient objects that are lost to us forever.

tl;dr - trying to make qualitative assessments of ancient technological advancements is incredibly difficult, both due to our own inherent biases as modern humans as well as the 'survival bias' of the artefacts and knowledge that have survived the erosion of time.

Among those who study innovations and technology, there is strong support and possibly an emerging consensus for the view that technologies are recombinations of pre-existing "components", meaning both physical components and knowledge/skill required to harness what Arthur (2009) calls "phenomena", that is, natural effects that exist independently of humans and technology (e.g. electricity).

​

This results to what Arthur calls "combinatorial evolution" of technology, and in my opinion it explains reasonably well why we see long periods (millions of years) of little development and then, relatively suddenly, an explosion of new "designs." This is the typical shape of a curve of combinatorial evolution: once the number of available components reaches a critical mass, there is an explosion in the quantity of potential recombinations.

​

This makes intuitive sense: new technologies have to be realised by a combination of existing technologies. For example, steam engine requires good enough understanding of metallurgy to build the components required: one reason why the ancients couldn't have built a workable steam engine was because the skills required to reliably and cheaply manufacture cylinders to tight enough tolerances was derived from cannon making, the demands of which had spurred the development of new methods for boring large, accurate holes in metal objects. This in turn was dependent on gunpowder, since before gunpowder there was obviously little need for cannon.

​

If you are interested in the topic, I can heartily recommend Arthur's 2009 book, *The Nature of Technology*. It is in my opinion the best single book for explaining the inner workings of the phenomenon we call technology, even though it largely ignores the social context (for that there are other books though).

There were many however this question is incredibly broad technological innovations in what exactly warfare? Metallurgy? Medicines? Agriculture? Astronomy? Not to mention the timeframe you have given many creations and advancements have either been lost or appropriated and expanded on by other later civilisations. Also consider the priorities and perspectives would differ from one continent to another learning more efficient irrigation and crop growing techniques in a desert like civilisation is great but would be less useful where crops grew well but they were more prone to invasion.