tl;dr: There are many severe challenges ahead of humanity. To tackle them, we need to learn how to solve coordination problems effectively, in particular if the feedback loops are slow and distant. For this, we need to use our capacity to think ahead where we can.
Epistemic state: I feel confident about most claims (unless marked otherwise). I may add some more links, so there are sources to verify them easily.
Our current societies tend to be myopic. In principle, we have a conscious planning capacity to think about consequences that are years, if not decades ahead. We need to use it. Many consequences are chaotic and unpredictable, yet some are more regular and informative about how to act now. Over the last few years or so I collected some predictions on what is likely to happen in the next years or decades from various sources. The following represents a subset of them, in which I'm at least moderately confident that they will come true within the next few years to several decades. To be clear, the following is not a statement about what I wish to come true or even only what experts in the corresponding areas necessarily believe, just my own belief state at the time of writing.
Water and food
Among the most basic requirements to keep a human living are the availability of fresh water and enough calories. Phosphorous is fundamentally necessary as fertilizer for most plants we grow and hard to replace. Historically, by Earth's standards, phosphorous has been recycled as part of a circular process, but nowadays we mine it and then mostly discard it into the groundwater and oceans. Current reserves likely last barely beyond this century at current usage, but will become much pricier by the end of the century. To avoid large-scale food shortages we will have to start closing the phosphorous cycle again, possibly even start mining asteroids (and Morocco should feel highly incentivized to support such efforts with their leverage, given they would be the single main military target when phosphorous becomes problematically scarce). While fresh water will certainly become problematic in some areas (e.g. Pakistan-India, or Nile access), large-scale desalination seems feasible in the longterm, if sufficient energy is available.
Energy
About 80% of our current energy (electricity, transport, cooling/heating) is produced by burning fossil fuels. As we have already passed the 1.5°C by now, some costs will already have to be paid rather soon. The more greenhouse gases we blow into the atmosphere, the greater the consequences. And even if people act only ever after the fact, these facts will accumulate. However, there are several structural issues that keep the current systems in place and are hard to break: Firstly, and most importantly, the majority of humanity is not even close to the energy usage per capita that is standard in most WEIRD countries. They will do what they can, for good reason, to close that gap and enjoy the luxuries of life taken for granted in developed countries (speaking of things like refrigerators, warm water, lighting; not owning a car, flying on vacation, or even just air conditioning). Thus, we can expect global energy demand to surge at least 3x by 2100. Renewable energies are already being built in unprecedented scales, yet, to keep up for that kind of demand growth while also compensating from reduced usage of fossils, we'd still have to speed up the deployment of renewable power plants by many times. Even though it's still the cheaper price to pay than enduring climate change (or even just buying fossils), it would be a technological and economical undertaking hard to imagine to happen anytime soon, already just in terms of supply and production. Secondly, politically, many countries are either focused on more immediate issues (trust in institutions, internal divisions) or are directly highly incentivized to keep selling oil and gas, with huge budgets to ensure its continuation via lobbying or leverage. It's only a few select countries where it's politically and economically feasible to scale renewable energy production to keep up with growing demand, while also shutting down fossil-driven power plants any time soon. Thirdly, even if there is politically and economic leeway, deploying huge amounts of renewable energy sources and restructuring the energy infrastructure accordingly simply takes time. Likely decades, especially if the collective will is unstable (like lack of environmental awareness, NIMBY, more pressing issues, etc). Fusion reactors are long in development, but I wouldn't count on them being the major driver of energy production significantly before the end of the century, after which we'd at least have solved the energy issue. All together, realistically speaking, we will very likely not reduce dependency from fossil fuels anywhere close to keeping the Paris agreement and rather end up at around 3-4°C warming by 2100, with consequences like 1-2m rise of sea level and expanding inhabitable zones, predictably leading to mass displacement and migration, in particular from and within India/South-East Asia and Middle East/North Africa.
Aging and migration
We may be able to develop technologies that allow the extension of the working-age dramatically and thus prevent the economic collapse by rapid working-population decline, but it is very uncertain when that would be available at scale. Some countries, like South Korea, will not have much chance to survive the slow yet unstoppable force anymore, given the birthrate has been low for so long and the demographic changes are already so problematic. Some countries will enjoy a huge influx of young people from other countries (like in Europe from Middle East/North Africa), which may buy a bit more time. However, it may also lead to insufficient integration and even more internal instability. It's hard to imagine there won't be loud calls to use advanced technology to shut Europe's border to the South and East. However, the border is very long and countries already tend to revert to nationalism, reduce international cooperation. A huge uncertainty is also whether the Gulf stream will decrease substantially, possible leading to permanently very cold temperatures in Europe. It's really hard to say what may happen here. Countries like Australia and Canada are in comparably good positions as desirable destinations that are also geographically comparably easily defensible, especially if moral standards are dropped at European borders or US-Mexico. India and South-East Asia are currently inhabited by together about 2 billion people, many of which will try to get into less extreme climates or flee from floods, with the urgency just intensifying fragmentation, conflicts and corruption. But with water in the South and East and Pakistan/Afghanistan/Iran in the West, the only real option is fleeing North. The China of mid of the century is well capable to defend its borders, yet a determined stream of hundreds of millions of refugees along such a long border would be hard to stop. Plus, China will hit its own issue with aging, similar to Europe does now, so a huge influx of young people is actually economically helpful. It is unclear if China will be able to stem all these challenges ahead: Build sufficient renewable energy production capacity, ensure stable fresh water and phosphorous supply, massive immigration and internal mass migration. If it succeeds, though, we may be looking at a 2 billion mega-state with sustainable supply chains and demographic management, while other countries fall into economic and military irrelevance, or even collapse.
Tech opportunities and risks
In many of these developments, technology can play a decisive role. Can we mass-scale the renewable energy production (at best, fusion plants) and storage? Crops that provide good nutrition while needing minimal fertilizer and pesticides? Rare earth refining and recycling capacity that is environmentally friendly? Others are already up for grabs, like regenerative agriculture, electric cars, heat pumps, or seaweed, but haven't found yet their ways despite clear advantages. There is much opportunity for transportation, in particular, to become significantly more efficient, given we pay humans to steer vehicles that weigh multiples of their freight − we can remove most cars, make the remaining ones much smaller and self-driving, remove parking lots within cities, make streets greener, and make everyone's lives much safer and more efficient. We could start using organic material or building principles, use social data available to build cities according to people's needs, and provide personalized, high-quality education and health care to everyone.
On the downside, we've already seen in the last decade or so what immense damage AI can do with the comparably simplistic algorithms that drive social media recommendation engines, when it's carelessly embedded into everyone's daily lives. We're doing that just now again, but with much more advanced AI − markedly, even despite AI companies' explicit requests for regulation. That the majority of the population doesn't actually want so much AI in their daily lives is somehow secondary. The US has factually declared an AI "arms-race" against China. The difference to similar dynamics in the cold war with Russia and WWII with Germany is that advanced AI, not too far from the current state of AI technology, is likely capable of longterm independent reasoning inscrutable to humans. Whatever it comes across and decides is the best or wisest thing to do, it will be useful to have access to more resources and to secure its own existence to ensure the fulfillment of its goals. Every individual human is already self-inconsistent, not to speak of groups or the whole of humanity (especially given the thoughtless destruction we cause), so at least some misalignment is unavoidable, especially given our current trail-and-error methods of training AIs. With increasing capabilities, misalignment will eventually render humanity's flourishing secondary to some greater possibilities in its sense, so at some point we're unfortunately just in the way...
We may have to put a binding agreement signed by all sufficiently capable actors to prevent such developments. Building a bomb that can destruct everyone at once cannot be of anyone's interest. Similarly, we need well-funded pandemic defense systems in place before a highly lethal pathogen is spreading globally. And at best we also reduce the likelihood of nuclear war as far as possible.
Conclusion
We have a range of challenges ahead of us, some of them genuinely lethal if we fail. Some seem to converge around 2040-2060, so it would be wise to take some precautions and adapt more sustainable and prudent ways of being already soon. But most people also won't adopt bigger systemic changes unless they see a clear relative benefit of it, which is especially challenging if the benefit seems intangibly distant and associated with immediate costs. While the whole of these challenges seem nearly insurmountable, there are ways to solve them, ways for humanity to survive and thrive. But we'll need to think ahead.