These are purely hypothetical, not definite!
Many are genuinely being researched, but this will only provide shallow insight. There are also many other factors that may hamper or inhibit these efforts outright.
Example of the function of an artificial womb that has been researched for a lamb.
I am certain that you are likely well aware of what CRISPR is by now, but if you aren't, it is essentially singling out a specific gene that may need to be edited or appended. Once that gene is found, the DNA is unraveled and broken to then replace base pairs or insert a new string of DNA, thus creating a desired outcome. For this form of theoretical artificial human intelligence, genetic editing and engineering will be utilized in order to increase all aspects of the brain equating to increased intellectual output. If you were an astute viewer who read about some of the key genes necessary for human brain development, you would probably think that we could enhance or amplify the effects of those genes in order to achieve higher gyrification of the surface, higher total neocortex neuron count, and other important factors. Now clever you might think, "won't the brain get squished inside the skull then?" The answer is yes, so the endocranial capacity would need to expand in order to fit the brain's expansions. Now another issue comes up. Wouldn't it be pretty much impossible to be birthed if the skull is too large? Once again, yes. To solve that, the embryo would be placed with an artificial womb to incubate in a simulated environment without the limitation of a skull size being present. Other issues may arise simply due to the relation of the brain to the rest of the body such as energy intake, support of the larger head, and other possible head functions needing a redesign to suit the new skull. This remains ultimately the most "human" option of the other choices as it is all purely biological.
Intelligence in a dish sounds dystopian, but it's already at our doorstep. If that is too eerie to imagine, compare it to your computer. There are many past and ongoing studies regarding cultivating a brain. So far, they have been for comparison against computers such of that belonging to data centers as human brains are incredibly efficient in the energy department and other niche computational categories. What makes a brain so applicable to this topic is its potential for massive sizes due to no bodily limitations. If purely provided support from machinery to maintain life, it could theoretically function. Now the question is, are we actually able to support an organoid solely from the machinery provided to support its biological needs, is up for debate. Hypothetically if that objective was acquired, that brain organoid could very well be provided mechanical means of interacting with the world. Imagine bearing witness to a robot walking around, yet it has the exact same intellectual functions as you.
Abbreviated comparison of a super computer and the Human Brain
If you've ever seen The Matrix, you probably know the scene in which Neo learns how to fight and how to fly a helicopter almost instantly. By interfacing with a computer through your brain, several biological bottlenecks are able to be subverted with the help of an external computer. Imagine the computer as an auxiliary brain that can store information, provide information, or even interact with digital devices or connections. You have likely heard of or seen Neuralink in which they implant chips within a handicapped recipient's brain that allows them to interact with digital devices such as playing a video game, typing, and web browsing. Now imagine this on a personal scale that allowed for memories to be stored at crystal quality, to retrieve anything from data that exists on internet, or learn something incredibly quickly. While this option definitely doesn't increase the base level of "intelligence", it opens up the door for many utilities for knowledge.
This last theoretical, as well as the most ambiguous, may be interpreted in a multitude of fashions that may or may not be applicable in reality. The first option is to have a multitude of people that have an implant that allows them to interface with other people that also have the implant. Think of it as a sort of telepathy where people can lend each other their computational intelligence. The electrical activity of one person's brain could be interpreted by another. If one person is struggling to solve a problem for say picturing a freshly picked and glistening apple spinning at three rotations per second, they could recruit other people's visual cortexes and have them improve the image until it is photo realistic before "sending" it back to the original person. Another application is essentially an alternative to the single large organoid intelligence. Having a multitude of organoid intelligence brains that can interface with each other and perform greater tasks that require more computational power. You could have millions of small organoids all linked together acting as specialized neurons for different types of computational logic.
Brain-to-brain interface (BTBI) for information transfer between human subjects. The emitter subject is shown on the left, where sensorimotor cortex activity was recorded using EEG electrodes. The emitter performed an imagery based binary motor task: imagery of the feet (bit value 0) versus imagery of the hands (bit value 1). The receiver subject is shown on the right. The TMS coil was positioned differently over the visual cortex for 1 and 0 bit values, and evoked or did not evoke phosphenes (flashes of light), respectively. An Internet link was used for this brain-to-brain communication.