Breeding with the help of AI (HAL420-Style)

Hi guys,
just out of curiosity: are you using AI like ChatGTP or Gemini for breeding projects?

I find it really helpful. I used it for my Kubricks Gold (see below) and the AI was quite on point. It also helps with F2, Reversals and stuff.

I crossed a female Malawi Gold with an iranian Landrace male (all seeds therealseedscompany). Thoughts?

This is a very interesting cross, and it’s exciting that you’re working with pure landrace genetics. Here’s a breakdown of what you can expect from crossing a female Malawi Gold with an Iranian Landrace male, based on the known characteristics of these strains.

The Parents: A Tale of Two Landraces

Malawi Gold (Female):

  • Genetics: This is a pure sativa landrace from Malawi in southeastern Africa.*

  • Effects: Known for a powerful, energetic, and cerebral high. It’s often described as uplifting, euphoric, and can even have psychedelic or mind-bending effects. It’s a great daytime strain for creativity and motivation.*

  • Growth: Malawi Gold is a tall, branchy plant with a long flowering period (up to 16 weeks). It’s best suited for outdoor grows and requires patience.*

  • Flavor/Aroma: Tends to be spicy, earthy, and citrusy, with some people reporting undertones of tropical fruit like banana.*

Iranian Landrace (Male):

  • Genetics: An Iranian landrace strain is a pure indica. The specific lineage can vary, but these strains have evolved over centuries to be well-adapted to their environment. “Mag Landrace” is one well-known example.*

  • Effects: The high from Iranian landrace strains is typically described as deeply relaxing and sedating. It’s often used for pain relief, insomnia, and stress. It can also have a euphoric, “giggly” onset before the body high takes over.*

  • Growth: These are classic indica plants: short, bushy, and with a relatively quick flowering time. They are known for high resin production and dense buds.*

  • Flavor/Aroma: Common terpene profiles include earthy, spicy, and even diesel or chemical notes.*

The Expected Offspring (F1 Generation)

When you cross a pure sativa with a pure indica, you get a hybrid. The offspring (F1 generation) will express a mix of traits from both parents. Since you’re using a male to pollinate, the male’s genetics will primarily contribute to the plant’s structure, flowering time, and overall vigor, while the female provides the foundation for the cannabinoid and terpene profile.

Here’s what you can generally anticipate:

  • Hybrid Vigor: This cross is likely to show significant hybrid vigor, meaning the plants will be robust, healthy, and potentially more resistant to pests and diseases than either parent alone.*

  • Structure: The plants will likely be a hybrid of the tall sativa and the bushy indica. Expect a plant that has some of the height and stretch of the Malawi Gold, but with a more compact and sturdy structure inherited from the Iranian male.*

  • Flowering Time: This is a key area of improvement. The long flowering time of the Malawi Gold (14-16 weeks) will almost certainly be shortened by the Iranian indica. You can probably expect a flowering time of around 10-12 weeks, which makes it much more manageable for most growers.*

  • Effects: The high will be a blend of both. You can expect the initial energetic and euphoric rush from the Malawi Gold, followed by a more relaxing and physical body high from the Iranian Landrace. This could create a very well-balanced effect, offering both mental clarity and physical relaxation.*

  • Aroma and Flavor: The terpene profiles will be a complex mix of the two parents. Look for combinations of spicy, earthy, citrus, and floral notes. There could be some truly unique and intriguing flavors.*

Thoughts and Advice for Growing

  • Phenotype Hunting: The beauty of working with F1 landrace crosses is that you will get a diverse range of phenotypes. This is your chance to “pheno hunt” and find a plant that expresses the specific traits you desire. You might find some that lean more sativa, some that are more indica-dominant, and some that are a perfect 50/50 blend.*

  • Breeding Potential: If you find a desirable female phenotype from this cross, you can keep it as a mother plant and either self-pollinate it with a male from the same cross to create F2 seeds, or you can cross it with another strain to create something new.*

  • Cultivation: Be prepared for some variation in your grow space. Some plants might stretch more than others. You might need to use techniques like topping or LST (Low Stress Training) to manage the height.*

  • Patience is a Virtue: While the flowering time will be shorter than a pure Malawi Gold, it will still be on the longer side for modern hybrids. Give the plants the time they need to mature and develop their full potential.*

In conclusion, this is a fantastic and intelligent cross. You are combining two very different, but very stable, landrace lines. The resulting plants should be vigorous, and you’ll likely get some truly special and unique phenotypes with a balanced, powerful high. This is exactly the kind of work that helps preserve and diversify the cannabis gene pool. Enjoy the process!

8 Likes

Interesting responses from simple inputs.
Accuracy TBD. :wink:

9 Likes

i use various types of AI at work.
what exactly are you having the AI do here?

I could see feeding it some table data or something and having it sort stats, but that can be done by hand pretty easily. Any ML process will be a pretty slow churn.

I thought it would have been better for problem identification…which it seems it is good for as commercial croppers are using it now from drones.

3 Likes

It is not good enough for this yet. It doesn’t know enough about cannabis. Its sources are rollitup, overgrow etc. The Iranian for example is not sedating or used for insomnia but is described as refreshing, relaxing and mind opener from anyone who has used it. You wouldn’t know that unless you knew people who grew it.

There is someone on my discord who grew it and there’s some Jade Nectar videos from right now where he describes it and specifically says that these strains are not sedating.

It is like listening to your overly confident friend who is always wrong, trying to explain something. Eventually you know not to listen to them but at first they sound so convincing.

6 Likes

Well, it helps me to decide what strains i might try to cross, which european seedbanks have it in stock, analyze the name i choose for the cross and telling me if a strain might be to dominant. A lot of things can be done. In case of my posted example it was pretty accurate.

Well, i grew a full pack of the iranian (Iranian – The Real Seed Company) four or five years ago indoor. It was sedating like a pure indica. Of course this is my personal experience. Maybe i got lucky with the phenos…

No, the sources are not only sites like overgrow. AI is way more clever than that.
Don’t take it the wrong way or think i am a smartass but I worked for over 25 years in digital publishing, web-design, project management and as a CDO. In the last 4 years AI was a huge part of my professional life and it improved in a really frightening speed.

I am not listening to an overly confident friend. It is a tool. And it can be a useful tool but will still have to do dirt digging by yourself. :nerd_face:

I know enough growers who are growing their stuff this or that way because of anecdotal evidence like white ash … some AI could help there too :laughing:

3 Likes

Well, in my case it was close to the result. :raised_hands:

I asked ChatGPT for its sources 2 years ago before it stopped listing them and it told me it got the information from cannabis forums.

1 Like

Yes, i feel you. Some stuff is old but it also depends which Versions you use. Every Site that is listed on google and doing SEO is crawled. As i said AI is evolving in lightspeed.

I am not here to convince you of anything or preaching some new method. Your opinion is your opinion and i am fine with it. As i said i was curious if someone uses it at Overgrow too.

2 Likes

It isnt an opinion. The technology isn’t reliable enough yet for things that barely have information. You can ask it about political history of Italy and it knows that really well but there isn’t a long history of Cannabis literature for it to take from so it has to scrape icmag, rollitup etc for information. Modern cannabis research hasn’t even started until recently due to extensive worldwide prohibition.

4 Likes

I believe @LonelyOC is working on a cannabis specific AI, that I have not had the chance to try yet due to being swamped. :sob::sweat_smile:

5 Likes

@AdvancedGenetics uses AI as well

1 Like

I dont understand what you need AI to tell you there that you couldnt figure out yourself…seems redundant…but i guess its quicker than reading through the forums yourself.

6 Likes

If u say so.

BTW u can AI also use for design, logo and stuff… just three examples.



1 Like

Sure. I am lazy.

2 Likes

Yeah, gave it up though. Couldn’t get any payment service to work with the site. They all denied me because of cannabis related content…

Unfortunately :weary:

Pz :v:t2:

3 Likes

Have you tried https://ko-fi.com/ ?

1 Like

Is it still in the making to make cannabis a scheduling 3 drug? might change the ballgame if it is really happening.

1 Like

No, I will take a look at it though! Thank for the heads up! :pray:t2:

The idea I created is that you have two trained AIs. One on data from icmag, OG, rollitup, podcasts, seed catalogs, books, pretty much everything I’ve been able to scrape and get my hands on. Then a wiki with curated content, and then a AI to that curated content. That makes it possible to overtime work together as a community to make the most accurate library of cannabis information that is searchable by AI.

Pz :v:t2:

6 Likes