# Could agriculture still be developed by intelligent species that lack the concept of past and future?

Imagine somewhere in the universe there is an intelligent species that is as intelligent as our stone age fore-brothers and sisters but will never establish the concept of past and future events.

They won't remember any history nor are they be able to do divination/prediction such as no weather forecast etc. Our hunter gathers used to hoard seeds and water in case of emergency which could happen in the future based on some tragedies such as long period of drought and famine that occurred in the past.

I am wondering how could such a species farm anything if lack the capability to compare any 2 events that may happen either simultaneously or across different times to develop agriculture?

• How can a species be intelligent without having the perception of time? – L.Dutch - Reinstate Monica May 4 '20 at 7:44
• Might want to take it back a step and ask how it could be and they still be able to survive, maybe "what alternative strategies/abilities might have evolved?" or something along those lines. – Tantalus' touch. May 4 '20 at 8:08
• @SZCZERZOKŁY How do you know you filled it 'today'? What even is the concept of 'filled'? Can you even think if you can't perceive change? You don't even have the vocabulary to order your thoughts, much less writing. – user3082 May 4 '20 at 12:24
• @SZCZERZOKŁY how are you making a bottle if you have no understanding of sequential activities. – John May 4 '20 at 12:34
• I'm going to challenge the premise here a bit. Any species which can't understand the concept of a past or future isn't intelligent, because it can't learn. There's no realistic way for a species to reason about the future without it having such a concept to begin with, and without that kind of reasoning, you don't have intelligence, just some biological analogue to an expert system. – Austin Hemmelgarn May 4 '20 at 16:34

If they have no sense of past or future, they live in a continuous present, answering the stimuli they are getting right now: pain? move away! hunger? eat! Thirst? drink. Nothing more than an amoeba.

I highly doubt that such a species could satisfy any definition of intelligence, let alone developing something as complex as agriculture. Don't forget that homo sapiens has been intelligent (about 130000 years) for way longer than it has been a farmer (about 20000 years).

Offering a dissenting opinion on those that say the concept of time or even only the perception of causal relationship is necessary to develop habits that show traits common with agriculture (or other "long term thinking" activities) the humans engage in.

First at all, a couple of definitions on my choice of terms (to address some points raised - rightfully - in the comments):

• agriculture - "The science or practice of farming, including cultivation of the soil for the growing of crops and the rearing of animals to provide food, wool, and other products."

• practice - "The customary, habitual, or expected procedure or way of doing of something."

Nothing in the above require planning, or the use of technology or a sense of time. Granted, the chance of doing agriculture right without them is small... but I'll endeavor to show that is not impossible.

I am wondering how could such a species farm anything if lack the capability to compare any 2 events that may happen either simultaneously or across different times to develop agriculture?

In the same way on which:

That way? By natural selection.

No, neither agriculture nor sanitation need to be rooted into an economic (or other "rational") reason to be practiced - survival is enough.

Now, looking to the issue of intelligence: the examples above do show unintelligent (by human standards) species engaging in activities that humans do "intelligently".

Suppose such a species which ends (by evolutionary pressures) in adopting "long-term-effects type of behaviour". Because of this, their individuals grow stronger and/or live longer and/or are able to direct the extra energy into something else. Also assume that the species evolved a brain over a critical mass.
Is it impossible that the intelligence they develop to come as an effect of the species "habits", rather than intelligence being the cause of their habits? (e.g. farming fungus and having the physiological needs secured free the time/energy that lead them to develop and use tools)

Speaking of intelligence/reason/rationality - humans are expected to be this way, right? Then explain to me the irrationality of the stock market

• just a note squirrels and mice definitely have a concept of time. and there is a difference between farming and agriculture. Agriculture is a technology, a planned action. – John May 4 '20 at 14:55
• You are right under the definition of the science ... of farming, including cultivation of the soil for the growing of crops and the rearing of animals to provide food, wool, and other products. However the Oxford dictionary uses: The science **or practice** of farming.... Which allows the interpretation of "a set of practices related to farming is still agriculture" – Adrian Colomitchi May 4 '20 at 15:09
• Hence why I clarified my use of the term, I am using agriculture as the technological application of farming. clarity is important is a discussion like this to differentiate what an ant does to what a human does. What word would you use to specify "the purposeful planned activity of farming", replace agriculture with that term if it makes it easier to understand. – John May 4 '20 at 15:30
• Ants (at least around here) also plant flora, as well. However, ants also know that they are seeing a reflection of themselves in a mirror, which many other animals are not. So... measuring intelligence is not as easy as might be imagined. – user3082 May 5 '20 at 4:55
• @awelotta I suggest that a possible way for the hypothetical creature reached the stage of developed intelligence AND incapable of perceiving time is by evolving farming while they were unintelligent, which helped them grow a bigger brain and let them time to exercise it . Because the question is ... how could such a species farm anything if lack the capability to compare any 2 events that may happen either simultaneously or across different times to develop agriculture? and not given an intelligent species with no time perception and no agriculture, can it develop the latter? – Adrian Colomitchi May 5 '20 at 22:43

Hyper-instinctual savants:

Your beings would have to be skirting the limits of what WE, at least, define as intelligent. It would mean that the individuals would need to be performing complex tasks out of an instinctual drive, but still be intelligent. The species would be VERY rigid in their ability to respond to problems. My answer would not meet the criteria most people would consider, but it's the best I can envision.

Your aliens are essentially performing a series of problem solving tasks that they abstract from their instinct. PROBLEM: flint absence. SOLUTION: flint-seeking (see clay subroutine, access memory) MEMORY: Clay deposit with flint. PROBLEM: seeking behavior for flint SOLUTION: walking to the clay bed. (see clay bed memory). Ooh! flint! (see pick up subroutine) PROBLEM: I am not in the village. SOLUTION: walk to village. MEMORY: village location. PROBLEM: I am carrying flint. SOLUTION: make a tool. PROBLEM: I am making a tool from flint. SOLUTION: access tool prioritization subroutine.

You would have the equivalent of brilliant problem solvers who wouldn't know why they were solving the problems. They might abstract cause-and-effect, but only to answer questions, not because they cared about the outcome or understood what they were doing. They would look a lot like the clever ants we were discussing.

Not everyone would define this as sentient, but I know people who don't think extremely autistic people or even babies are sentient. Computers would function a lot like this, and perhaps this might be what machine intelligence could look like.

• But memory implies having experience from the past. With no sense of the past it would be just the proverbial gold fish constantly experiencing new things even when going around in a bowl – L.Dutch - Reinstate Monica May 4 '20 at 16:16
• Is it plausible that they have knowledge without an understanding of how they came to have that knowledge? I think the OP might need to weigh in whether that counts as a sense of time or not. – Matthew May 4 '20 at 17:15
• Hey, folks, I couldn't come up with anything that worked if you excluded memory. I have to settle with lack of awareness of the past. – DWKraus May 4 '20 at 17:43
• You can get something like memory but that doesn't require a conscious process by moving items to specific places when certain conditions are currently met, and you can get rather complex behaviours from evaluating the current state of the items, also on a continuous basis. Plants have rather complex behaviour patterns depending on temperature and direction of sunlight, but don't require neurological processes for that -- chemical signaling is largely sufficient. – Simon Richter May 4 '20 at 19:41
• I would see an interesting opportunity to have a unique programming/educational system here. It would be tricky to work out, but this species could serve as a substitute for computers/slave race, who 'programs' the members by giving them books with memorizable subroutines. How they would reach that level would be the trick. The individuals would act like machines, but SOMETHING (fill in the blank) runs the show. These individuals would be better adapted to specialized roles in an advanced, industrialized society than a primitive existence. – DWKraus May 5 '20 at 0:34

Offering an alternative...

Most answers consider they lack a concept of past/future because they just haven't developed it. To me, it seems vital for development as we know it.

However, there's another alternative: They have no need of the concept of time. Consider if they, as a species, were unbound from the flow of time.

Past and future would look very different to them. It wouldn't be thinking about the future, it'd be reverse-engineering the future to get the desired result. They might not know why spreading manure helps plants grow, but they actively see the results of their actions as they do them, so farming is less like planning and more like painting and tweaking until you're looking at the desired results or drawing with a spiralgraph.

Depending on how time works in your setting, they may even have a bi-directional relationship with time, and they may even do things in the future to change the past.

• Tralfamadorianism, back at it again. – Fivesideddice May 5 '20 at 12:21
• Was about to post an answer on this track. As I imagined it, they could traverse time as freely as we traverse our 3 spatial dimensions (or at least 2 of them, because gravity). From how a quantum physicist understands "past" and "future", the words might be applicable. But the stone-age concepts of "past" and "future" surely don't apply, it's more like "here" and "there". For these creatures, a seed "flows" (transforms) into a plant like for us water flows downhill (and can go uphill, if we push it). – Blueriver May 5 '20 at 16:52

## Stigmergy

Lack of the concept of past and future essentially turns your species into operators of finite-state machines. While FSMs have limitations, you can do quite a lot with them. See this article from an example in real life.

Say your intelligent aliens want to water their crops every 10 days. Here is a simple scheme that they can use without requiring any internal memory: (by "memory" I mean writable memory; they need to remember something to have intelligence)

• If there are stones next to a crop, take one of them, throw it away, and move on to the next crop;
• If there are no stones, water the crop, put 9 stones next to it, and move on to the next crop.

You can add more rules, such as "if it's rainy then put 9 stones next to the crops and move on", etc. The point is to move the memory requirement out of the aliens to some external representation of "state", like stones in this example which signal the number of days before the next watering.

How could they have developed such a complex scheme, you ask? Natural selection would be an answer: each individual mutates their operating rules of stepping the FSM a little bit (maybe putting 8 or 10 stones would be better than 9 stones?), and only those works can survive.

• 'If it is rainy, add stones until they number 9 and move on". Otherwise you can have 17 stones, given bad timing. And how much rain is 'enough' - you will need a rain-gauge (bowl?) or something as well. – user3082 May 5 '20 at 4:44
• And how much rain is 'enough' - until the evolved biological sensor says is enough. If the sensed level is wrong, the organism/colony will die while others with better tuned values will survive. It's selection, don't you dare go accusing nature of intelligence! – Adrian Colomitchi May 5 '20 at 5:48
• Isn't it proven that a Turing machine can run any program, including an AI? Not sure whether general-purpose AI is even possible, and I know a Turing machine requires an infinite tape. – Blueriver May 5 '20 at 16:57
• And "Fred the human" comes in, picks up a rock as a souvenir, and throws the entire society into disarray. – lilHar May 6 '20 at 17:03

# My interpretation of the species

These are my assumptions for a plausible species that would classify as intelligent and verbal, but have no concept of past or future.

The creatures have brains that can process time, and form memories, at an autonomic/emotional level, but not a cognitive one. This means they have no reasoning, logical, or verbal capacity when it comes to memories, recurring events, and outcomes of their actions. Their cognitive abilities are otherwise normal, including learned facts that don't involve time, and abstract reasoning.

They can identify familiar places (perhaps even attach names to them), and have fears based on past experience (perhaps even visualising those bee stings), but cannot reason about any past facts or future outcomes.

For example, if they want to want to drink, they will feel a subconscious attraction to the river (and their brains have some innate sense of direction that doesn't require thinking, which I think nearly all animals have to some extent).

But they can't plan to avoid the beehive on the way. They can't plan to avoid the giant fallen tree.

To be clear, they can reason about "stimulus-response", just not "result". They can remember what they should do under certain conditions, and reason about the conditions and the actions. Just not the expected outcome.

# Development of knowledge

Knowledge and technology develop slowly. Elders can teach the young ones what herb to rub on a rash. This fact can be verbalised and remembered without any reference to past or future. But how was that knowledge found in the first place, given that when someone tried the herb, nobody remembered it when the rash got better after a few days (or even hours)?

The answer is that members of the species just do random things a lot. They don't remember what they've tried before. At a subconscious level they form "associations" with what went well and what didn't, and over time, given a situation they will "feel like" doing something. Slowly they become confident of the fact, "when abc, I do xyz". Crucially, there is no "... then efg will happen." Stimulus-response, but no consequence.

(If the herb isn't there, another fact is when [Rash + no herb] do [walk around and take the rashy person with you] so they don't forget what they are looking for.)

Young ones will then repeatedly put into practice the advice of elders, and over time form associations of which elders have the best advice. This ensures that mostly good advice is passed down generations. They can't remember who told them something, but their brain will form subconscious associations, because that's one of the most basic things animal brains do.

# Agriculture

This is really just the same thing at higher complexity. It must develop very incrementally. At first, facts like "if you have fruit pits, bury them", later, "if you see a fruit tree and it looks wilted, put water on it".

"At sunrise, do xyz" will ensure something happens once a day. (They can identify sunrise, just not predict it.)

"Twice a day" is meaningless (but can be simulated with the stone-system proposed by nalzok.)

You can do a lot with stimulus-response if you have sophisticated stimulus classification and if the responses can refer to parts of the stimulus. I think simple agriculture is definitely possible, and complex agriculture is plausible.

# Non Temporal Species

The only way I see this working is if the creature is a non temporal species kind of like the Prophets from Star Trek that do not experience time in a linear fashion like we do. For this species there is no concept of the past, present, or future since to them all of it is present tense and is currently happening. From the time they are born to the time they die they already know everything they will do or have done, and it cannot be changed.

## Predicting Future Events

They do not predict future events nor plan for them in the way we would, since to them the future is happening right now. They plant seeds because that is what is happening just as harvesting the crops is happening at the same time. They do not necessarily understand causality in the same way we do. For them they could be unharvesting the crop so that it would shrink into the ground to produce seeds. They would not necessarily be able to tell a creature experiencing time linearly which event happened first or whether the even happened before or after the event of them interacting with the creature.

## Language Construct

Since they would have no concept of past, present, or future their language also would lack these characteristics. Their conversations to us would sound very confusing since they would be talking about an event and it would be impossible for us to tell if they are talking about something in the past or future. Any literature they write would also not flow linearly and work under the assumption that you already read it.

## No and All History at the Same Time

History much like literature would be extremely no linear and impossible to parse. Much like how human history gets past down to the next generation they would be able to pass history up to the previous generations. This would result in history books being absolutely convoluted mix of past, present, and future events. As such to prevent massive volumes of gibberish from piling up they likely would not even bother recording it.

## Concept of Life and Death

Since they know their death as well as their birth, they are experiencing their entire life all at once, but since they have no concept of time their life is perceived as being eternal. To them their birth would be viewed in the same perspective as their death, it just represents the other end of their existence. As such they would not necessarily fear death nor be concerned about what would come after it in the same way we do not fear our birth or what would happen before it.

Just hope one of them does not greet you and says "Your end of existence celebration is lots of fun, and I am enjoying it greatly."

http://www.exactlywhatistime.com/philosophy-of-time/ancient-philosophy/ Concepts of time, as philosophy concept, postdate agriculture. Ie "will never establish the concept of past and future events."

I suspect you want even less cognitive ability, but I don't think you need to understand time to have an understanding of cause and effect outcome. I don't think a dog understands time, but understand getting a leash means walks. A chimp still builds a nest to sleep in.

Unless you want to say they have no memory of their action to associate with the effect. In which case I'm struggling to see a pathway forward.

• you absolutely need a concept of time to learn cause and effect, otherwise you would be perpetually confused as to why washing the dishes does not cure your hunger. dogs absolutely understand time, rats understand past vs future. – John May 4 '20 at 12:31
• Hard to check, but interesting in any case (at what point the "operant conditioning" by environment becomes complex enough be called intelligence). In any case, an inter-generational transfer of knowledge/experience must exist (not an uncommon behaviour) – Adrian Colomitchi May 4 '20 at 12:34
• @John Causality, learning patterns and the concept of past/future are different things. Maybe not totally orthogonal, but they are different. Things like cargo cults aren't very intelligent by our measure - a dog would never do anything like that; so, is a dog more intelligent than the cargo cult members? – Adrian Colomitchi May 4 '20 at 12:37
• A dog can catch a ball. It doesn't understand the trigonometry to calculate the future location of the ball. It knows the answer, but could not show you the working out. The dog does not have 'a concept of trigonometry" – Commander Nirvanah Crane May 4 '20 at 12:58
• @John you only need to demonstrate now that the practice of agriculture requires planning and problems solving and cannot be done in an instinctual way (evolved behaviour) – Adrian Colomitchi May 4 '20 at 15:12

For a being that cannot understand the concept of past and present , simply will not be able to develop agriculture.
It is because they cannot co-relate the event of spiting/burying/dropping a seed to sprouting of a plant.

Also a being which does not remember the past cannot collect knowledge and can never pass it on.For example though octopus have very high cognitive functionality they are not able to take over the world because they have a short life-span and do not stay around to pass on the knowledge to their off springs.So a being which has no memory and forethought about the future , will have to keep reinventing the wheel.

But still this being could leverage autotroph.That could be from a way of a symbiotic relationship , where a being capable of autotrophy gets inside of your species (by eating) or grows on top of your species. However the latter is unlikely.The advantage for the autotroph is that it could get access to its energy source(sunlight) , protection and also spreading.

Let's briefly review key points of the question.

Context of the question: There is a species, and it is intelligent. The species has no concept of past or future. What we want to know: How could this species farm something? EDIT: Asking this part actually requires that the species farm as part of the context, so I can't argue against their ability to farm - only about the way they could achieve it!

First thoughts: doubt about the species's supposed intelligence - They don't have a concept of time? How could they be intelligent? Unfortunately for me and likely others, following this line of thought would turn my answer into a strawman argument, since assuming the species is actually not intelligent would mean that I would actually answer some other question, and not the one at hand. Moving on.

Natural selection - Regardless of why the species does things, the better ones survive. This means that even if the species had no notable intelligence, and simply operated on instinct, the species would evolve in a way that was generally beneficial. Natural selection could shape this species into one that practices agriculture, even if unlikely. Their instincts alone could achieve this with the right circumstances. A community of the species that benefits its food source will survive better, even if they did not act based on intelligent thought. I feel that we can assume that, regardless of their intelligence, it is possible for the species to develop agricultural tendencies, since they don't necessarily have to be thinking about it.

If we were wondering how they could actively develop farming techniques in manner that was based specifically on one generation of the species and thus without evolutionary influence, my answer might be different.

In the context of world-building, they could be a species that is not well-understood by others because of the supposed paradox that they represent. After all, it's very counter-intuitive that they have farming practices when they don't have recollection or prediction of events. The truth could just be that they simply farm because its "hard-wired," and not because they decide to.

As for how you were to describe their intelligence without referencing ways they predict or recall? Can't help you there.

EDIT* Actually, you could describe their intelligence as an understanding of what things are... not as a mental construct, like "shovel," where it's an understanding of what abstract actions its used for, but an innate understanding of the exact structure of the item as it is in that instant. They would likely not be able to understand the concept of a shovel though. Not completely sure if that counts as intelligence.

I think that there'll be a bit of trouble with definitions, semantics, and philosophical aspects of the question. In my opinion, the very comparison of given species to the mentally-butchered form of our ancestors isn't perfect as a problem statement, since the created gap is so huge, that - at least to me, currently - it's hard to establish whether the hypothetical species would be closer to an animal, computer or some exotic (and probably desired by OP) form of an alien human-equivalent.

It's easy to make the question easier by altering the environment to simplify the process of agriculture. The only problem is, how many elements can we remove before defined the process stops being agriculture?

We can make the environment to always have perfect weather, natural irrigation, fertile lands... so on, but at some point, it might stop being less like agriculture and start to be more like a combination of gathering and incidental pollination (seeding) that evolved into its instinctive counterpart.

If I could remove the requirement of 'development' - which in itself is a problematic aspect to define, then for me, the easiest answer would be species with an in-built computer equivalent. Since it was never stated in a problem statement that species were to be developed naturally, we could simply apply the idea of a higher-intelligence who assists the species or engineered their biology, for example:

• Based on the currently observed state of the environment - (the more 'sensors', the better), species 'intuitively' perform actions determined by their biological or external circuit.
• Example A: Engineered tree-like species with wide roots that inspect the soil and search for water, with perhaps hairy leaves and numerous eyes that act as the above-ground sensors related to weather conditions or detecting pests. They could act in an energy-conserving mode for most of the year, then refill their biological battery (fats) at the time of (likely a very massive) harvest. Perhaps, they could also have a biological system that allows the preservation of food, like submerging collected fruits in resin - I'm not into chemistry, so I don't know how it could work, but well... as stated earlier, we can always cheat and make products of our agricultural work almost never rot (think super beans/rice).
• Example B: Your typical caveman carrying a computer/device that tells him what to do.

However, that doesn't answer the original question, which concerns the aspect of development. We could substitute the 'instant' engineering with a progressive one or teaching/conditioning over time, but does that count as 'the species developed agriculture'?

Develop: 'to (cause something to) grow or change into a more advanced, larger, or stronger form'.

To my understanding of the English, the species would cause agriculture to develop, but only indirectly, more like slaves than the mind behind the idea, but I guess it would still count by a technicality.

IMO, the better idea would be the evolutionary roadmap, which ensures independence, example:

Start: Beetle-like creatures who accidentally carry seeds between their spiky hairs, like pollen.

Goal: Digging holes/rows for seeds

Evolution step: Some beetles evolve a talon under their belly, which breaks up the soil as beetle moves. The advantage in the process of natural selection is the extra combat ability, ability quickly to burrow to hide from predators, a bigger chance for accidentally burrowed seeds to develop in the home area - more food. Of course, it expands a lot of energy to constantly fight friction/soil, so it can be substituted just by digging holes for protection and in that case, the talon could be replaced by a horn, like in case of rhinoceros beetles. Another benefit is burrowing eggs for protection.

Goal: Digging rows

Evolution step: Eggs burrowed separately can have a higher chance to survive/give offspring. For example, because due to varying circumstances, predators have to expend more energy to dig them out separately than it takes to burrow. Offspring isn't forced to compete for food in clusters, instead, each gets its own patch of land.

Goal: Irrigation

Evolution step: Beetles could naturally choose to live close to water, to avoids deaths related to dehydration. Water could also be related to the development of eggs, perhaps beetles could instead start as water or hybrid creature (think frogs). Accidental tunnels could turn into structures derived from natural selection, that facilitate breeding and allow parents to take better care of offspring. Of course, we also remember the default advantage of more food due to a better chance of seed sprouting.

Goal: Fertilizer

Evolution step: Could be like a byproduct of dung beetles and accidental fertilization (think substances like pollen/syrup/water plant residue/the current carrying dead water creatures to irrigation tunnels/traps like spiderwebs and hunting/killing more than necessary - also, might be not for food to not overwhelm agricultural aspects, but eg. prey might be inedible/poisonous and our species might be territorial). Maybe an idea with using crushed shells? Carrying these for protection, eg. like hermit crabs.

Goal: Pest Control

Evolution step: Default advantage + more food

Goal: Weed Control

Evolution step: Creatures may use plants for clothes, hives, or structures. Using exclusively weeds leads to default advantage.

So on... so on... luckily, it can always be altered by simplifying the environment.

Still, the question that arises again is - did the species 'develop' the agriculture? To my understanding of the English, due to technicality - it would be 'YES', but it's more like the main 'author' is the sole process of evolution and (again) species acted as 'slave' to the process.

I've kind of wanted to discuss the concept of creatures in relation to the Turing machines, where memory could be compared to the machine's tape or creatures could create 'notes/guides' (for themself, or future generations), but it's kinda late, so I can't elaborate on that and I've kinda messed some details due to time anyway, so just giving a general idea. The concept is to write the notes in such form, that there's no hint or intent of planning for future or extrapolating from the past, but rather something like a written communication of ideas/results between two creatures, which would later develop into a kind of social framework/dogma, with no relation to past/future, but rather as the establishment of (perhaps nigh religion-like) facts (commandments). There're some issues to discuss, as to how far can we take writing into memory before it counts as related to past/future issues, controversies with awareness - eg. DNA kinda establishes the concept of past/future events but creatures are not aware of it, the same can apply to other types of memory and you can, of course, bypass many issues by making the creatures, for example, deluded af.

The answer gets already accepted but I am obliged to disagree. First: is it true that time consciousness may appear as a requirement for intelligence. But intelligence is a really hard notion to define actually there are several definitions of intelligence IQ being only one of them. In my point of view, intelligence is the capacity to adapt to change in environment. This definition is as same as the other as its flaws. But if you get rid of what you expected intelligence to be and consider this definition then IMO is it actually possible to make some basic agriculture. How?

First let's compare to what exists, what species do you know that does agriculture except for man? ant. ant still have from what we know time perception. I doubt they are conscious about it but they perceive it. If your individual perceive time in the sense of being able to register an immediate change to their environment, which is IMO the bare minimum to be able to adapt to change then they may be able to make agriculture without realising it themself. We all have innate behaviour most common for human are basic need sleep, food, water, sex. But we are also innately good at living with others.

The point is that maybe they don't even know they doing agriculture but some of their innate behaviour make them do so, very basically. Let's imagine their brain when it starts to get warmer (spring) will make them search frenetically the soil (labour) then they tend to poop, or throw their waste there or other behaviour that will make them somehow seeds.

that what we will call agriculture, very basic one, low yields for sure. If they developed that behaviour and selected them, in an evolutionary sense, that may mean it works good enough for them. and the one that didn't have them disappears long ago ...

Basic Agriculture does not require intelligence, memory or a notion of time.

Natural selection is, by definition, entirely sufficient to encourage instinctive behaviours that increase the species chance of survival.

It isn't likely to look like what we think of as farming - specific areas designated for farming are much less likely. But for example, scattering seeds from plants in appropriate places, cutting back unhelpful plants, and chasing off pests are all entirely feasible.

The fact that they don't know why something is good, does not necessarily prevent them from doing those things, though it is likely to be a lot more haphazard in appearance.

Can they be intelligent?

Intelligence can be defined as extremely good pattern matching. Preventing that intelligence from considering patterns that involve temporal aspects does not prevent 'extremely good' pattern matching from arising, though it does significantly limit the number of avenues it has to 'prove' that it meats the qualification of 'extremely good'.

It is probable that their non-temporal pattern matching would be better than ours - much like how a blind person is often better than a sighted person with their other senses.

It is most unlikely, but it would generally be possible. Not being aware means that you have no notion of "memory" or of "cause and effect". But! But!

There exists genetic memory, and your species might very well have that. We do, so why wouldn't they.

For example, human males (basically all higher animals) know exactly what they are to do with a female, and females know exactly -- in accordance with some specific, individual parameters -- that they are to tolerate that. They know without someone telling or showing them. Why? Well, because those that don't function that way do not reproduce and die out.

Most animals fear thunder and lightning, and they fear fire. Why do they? There's no reason, is there? Who told them?
They fear it because those that don't have a reduced chance of getting old enough to reproduce. Avoiding instant death is a very sustainable strategy, with a high selectional advantage.

While it's rather unlikely, your primates could very well have that kind of genetic memory, only with farming. Why not. They wouldn't need to know the sequence of events (or have any sense of causality) for as long as they know "put seed in ground when sun stands there, and climate is like this" and "harvest when sun is standing there and climate is like that".

2020 and there still exist human tribes with no concept of agriculture and no idea on how to start a fire.

2020 and there still tribes of people living 1.7 million years in the past, just like monkeys.

So yeah, i strongly believe your creatures wouldn't do much better than those human tribes (sentinelese to name one)