Belum Lulus SMA, Kok Bisa Kuliah di Amerika?

Kenalan: kenapa dipanggil Fron Seneca 0:52

Victor: Hi everyone, welcome to episode zero zero two of Podcast KITA. Jadi ya hari ini kita punya Mas Fron.

Dan Mas Fron Seneca, tapi namanya bukan Fron Seneca. Nanti kita...

Fron: Bukan, bukan. Ya.

Victor: Jadi maybe we'll start first ya, Mas Fron. Maybe let's introduce yourself kan, namanya kok ada dua kan. Terus...

Fron: Boleh, boleh. Ya.

Victor: Terus sekarang di startup ngapain gitu.

Fron: Ya, ya. Ya, so pertama, you know, very happy ada di sini. Thank you for the invite. Yes, namaku sebenarnya nama aslinya Fahran, Fahran Kamili. Tapi dari SMP I think kayak Fahran, Fahran, Fahran tuh kayak dipelesetin lama-lama jadi Fahran, Fahran, Fron, right. So, I think banyak orang kira kayak Fron itu kayak Fahran di-Americanize.

But that's not the case. Soalnya dari SMP di Indo udah dipanggil Fron. So, that sticks. Terus kayak Fron Seneca, it's like Seneca itu is one of my favorite philosopher.

He's like a Roman philosopher.

1:58 Victor: Stoicism.

Fron: Stoicism. And then he also happens to also be like a very wealthy philosopher. So I think aku juga I think there is something about like interesting about like how Seneca it's like he is kind of wealthy tapi masih kayak pursuit, you know, higher ideals like philosophy.

So ya, I think and I think it's just like aku kayak decided Fron Seneca and then semua teman-temanku sekarang kayak, you know, knows me sebagai Fron Seneca. Terus lucunya adalah aku ada teman dekat di Texas.

Dia ngundang ke... Mereka couple ngundang aku ke nikahan mereka.

Terus kayak aku nyari namaku di tabling, table position kan. Nyari Fahran Kamili kok enggak ada. Terus kayak ternyata mereka, they know me as Fron Seneca the whole time I ketemu mereka gitu. Terus kayak, "Oh, Fron Seneca is there." Terus kayak and then in that wedding I finally told them, "You know my real name is not, bukan Fron Seneca." So ya, I think ya that's the branding that has been like nempel lah ya.

3:07 Victor: Ya, ya, ya. I wanna go a bit deeper into stoicism later.

Karena I think not many people di Indo has heard of Seneca or has heard of stoicism, but we can go a bit later.

Mimetic: iklan AI pakai maskot 3:19

Victor: Now I wanna talk about sekarang let's introduce your startup ya. Jadi-

Fron: Ya, boleh, boleh.

Victor: ...what does it do and like what do you do there and you know things like that.

Fron: Ya, so ya jadi sekarang aku sibuk building my own startup. Masih relatively baru banget sih. Kayak I quit my previous job just last month. Previous job also was working on a startup. I was one of the founder, actually. So, I quit that job to build my own startup. Tapi the previous job that I quit was in marketing and what I do now sekarang juga di marketing.

A bit different angle, I can tell you more about what is the angle exactly. Jadi kayak it is marketing, so I run ads. Aku run paid ads buat klien-klienku. And tapi-

Victor: Masih ini Lapis?

Fron: Bukan Lapis. Ya.

Victor: Kemarin yang Lapis.

4:19 Fron: Kemarin Lapis.

Victor: Ya kemarin kita ketemu kan.

Fron: Kemarin terakhir kita ketemu...

Victor: Masih Lapis.

Fron: Masih Lapis. Ya. Sekarang-

Victor: Oke, jadi sekarang lo udah quit Lapis.

Fron: Sekarang udah quit Lapis. Aku-

Victor: Do similar things.

Fron: Do similar things. Similar in a sense masih di marketing.

Namanya Mimetic.

Victor: Mimetic.

Fron: Mimetic. Ya.

So, di ada teori namanya Mimetic Theory. It's like teorinya adalah orang-orang tuh when they pick up a behavior, they mimicking something. Mimic something. So, that's where the name comes from Mimetic, Mimetic Theory. So, tapi I also yang what's unique in the approach that I'm, yang kucoba sekarang itu, tesisku adalah running marketing using kind of like a story-based kind of narrative.

So, what I mean by story-based is like kayak di story itu ada karakter. Ada karakter utama, ada supporting character. Terus kayak this character is kayak macam ini sih, macam kayak maskot sih jatuhnya.

And then so I design the maskot, the story, the character for each maskot. And then buat marketing material for my clients, I use those maskot to tell the story about my client's product or apa cerita tentang my client's, you know, value proposition. So that's the teori yang apa, the thesis that I'm trying to test in my current company.

5:51 Victor: Yang kemarin pas Lapis itu kan kita pakai AI to generate imagery-

Fron: That's right. That's right.

Victor: ...for iklan gitu kan.

Fron: Yes, yes.

Victor: Jadi bikin iklan pakai AI. Tapi sekarang pakai AI juga.

Fron: Pakai AI, yes.

Victor: Bikin maskot untuk bikin advertisement juga.

Fron: Yes, yes.

Victor: So, it's kind of like still-

Fron: Ya, so-

Victor: ...ads, but not really the paid ads.

Fron: It's still in the ad space.

Tapi kayak I realize I think dengan AI itu we can actually create very [kurang jelas] content. And then I think ada term kan AI slop. AI slop I think and banyak people yang push back AI because they think konten yang dibikin AI itu AI slop which I agree. Kadang and I think the problem it's like it's not the AI yang problematic. The problem menurutku adalah it's the creator who doesn't put like a soul into it.

Victor: The people.

6:51 Fron: The people in it. Ujung-ujungnya kayak AI menurutku kayak it's just another tool, right? I think I'm pretty sure zaman dulu kayak waktu kamera di invent, right?

Victor: Oh ya, absolutely. So much push back.

Fron: People think, "Oh, this is not real paintings." Yang benar itu kayak painting. And then I'm pretty sure waktu zaman Photoshop di introduce kayak, "Oh, this is not real photos."

Victor: In the old days, in the philosopher age, mereka kan pada ngomong-ngomong doang tapi enggak ada tertulis. Begitu tertulis sudah diciptakan, dia bilang, "Eh anak-anak muda ini pada baca doang, bukan mengingat."

Fron: Bukan mengingat.

Victor: Itu kan. Jadi there's always this push back.

Fron: So I think exactly. That's a very good point. I think AI, the push back against AI I think it's similar, it's a similar vibe lah, similar energy to that. It's a new technology and aku sebenarnya setuju kayak banyak konten AI yang emang jelek.

Victor: It's very bad.

Fron: Yes, very bad. Tapi it's not because of the AI.

Victor: It's because the people don't know how to do it properly.

Fron: It's because they don't know how to do it properly, exactly. Terus aku juga percaya there has to be like a heart and soul into it, right? You really need to create like a good story. You really need to try to understand what is going on, right? And then ujung-ujungnya kayak AI can be helpful tool buat bantu orang kan.

8:09 Victor: Coba ini walk through example ya. Misalnya kan Mas Fron Mimetic dapat klien misalnya Ferrari. Because I see you have a Ferrari hat ya. Jadi nyetor Ferrari ke sini loh.

Fron: Yes. This is my Ferrari hat. It's pretty iconic actually.

Victor: Jadi Ferrari datang, Mimetic tolong dong bantu kita. What is the methods and the ways that you would be helping them?

Fron: So prosesku kayak ujung-ujungnya marketing itu enggak repot kan sebenarnya. Marketing ujung-ujungnya it's telling about a story about my client to their target audience, right? So it's like two things. So first aku harus ngerti klienku tuh kayak what is their philosophy, right? Kayak Ferrari is such an iconic company, right? So I need to understand kind of what is Ferrari's history gitu kan. Terus kayak I need to understand apa aku harus ngerti kayak what makes Ferrari's car special gitu kan.

Terus and then basically really really understand klienku tuh kayak who they are, what they do, value proposition mereka itu apa gitu kan. Terus dari situ I come up with like kayak karakter gitu that represents the car gitu kan. So I guess like if I were kind of making like Ferrari's kind of character maybe I'm thinking about the horse, right? So maybe kind of make some character related to the horse, right?

Or kind of like another maybe the driver kind of like maybe I create characters yang ada helm F1 and then all that fancy suits of like an F1 driver gitu kan. So I think it really depends on the client dan juga sebenarnya apa yang mau dijual gitu kan. So I think mungkin kayak kalau Ferrari kan kemarin contoh mereka baru release their electric car for example, right? So maybe kind of like the character needs to represent that and also kind of like really ngerti kayak target marketnya itu siapa gitu kan.

So target marketnya itu kayak things like orang tua apa orang muda, banyakan cewek banyakan cowok gitu kan. Terus kayak what's their cultural background gitu kan. Jadi kayak I also fit the characters with the target audience as well gitu kan. Terus sama dengan AI juga kayak aku juga use AI to leverage to test a lot of multiple messages gitu angle gitu kan. Kayak mungkin ada satu angle yang nge-highlight desainnya, ada satu angle yang highlight mesinnya, ada angle yang nge-highlight I don't know the efficiency, right?

Terus dari itu bisa kelihatan tuh kayak and then also I specialize in term itu digital ads. So ads yang ada di Instagram, LinkedIn, Google, right? So karena semuanya digital aku bisa run also kind of like massive experiment, right? Trying things. Dari situ bisa kelihatan tuh kayak oh ini yang dapat kliknya banyak tuh ini gitu kan. Oh yang convertnya bagus tuh ini gitu kan. Let's double down on that atau let's kill the one that doesn't work, right?

So ya I think...

11:31 Victor: How is the startup doing now? Udah berapa lama ini?

Fron: So I guess yang ini tuh kayak it's only been, Mimetic only been a month.

Victor: Oh baru banget ya.

Fron: Baru banget. Mimetic has only been a month tapi kan like you said Lapis it's also been in the marketing. Itu tuh dari November. So I've been in the marketing field since last November which is kayak hampir setahun lah. Hampir setahun and itu juga we can zoom in into that. Sebenarnya itu pivot a little bit. Aku backgroundku kayak I guess my career kerjaanku kayak never really do marketing.

Aku selalu sebagai programmer. So kadang orang kaget kayak this is like a weird pivot a little bit from engineering like exact sciences to kind of like marketing which is tend to be more creative. Tapi di kepala aku kayak sebenernya ujung-ujungnya enggak ada bedanya sih.

Kenapa keluar dan jalan sendiri 12:31

Victor: I think this is very important karena kan with the rise of AI semua orang sekarang bisa bikin website, bisa bikin app. Dan sekarang banyak banget app sama website sama games tapi kurangnya apa marketing. Berarti kan mereka enggak bisa untuk mempopulerkan their creation kan. Jadi motivasinya kenapa nih Mas Fron? Kenapa kemarin sudah di Lapis kan, terus kenapa mau quit dan bikin baru sedangkan Lapis is also in marketing.

Fron: Ya, so I think aku orangnya tuh aku peduli I learn kayak I care a lot about creative freedom. I care a lot about creative freedom, right. And also I am more let's say risk toleran. In a sense kind of like I'm okay with failure gitu kan. And dan aku juga, so ya from that two things I realize I think you know it's better off untuk aku coba sendiri gitu kan.

So that I gain back my creative freedom. We can talk about, so sebelum Lapis aku sebenarnya ada juga bikin perusahaan sebelum itu. So aku udah ngerasain tuh rasanya kayak have like a complete creative freedom, right. Ada pros and cons-nya, we can talk about it. Tapi kayak I realize I really-really like that aspect of creative freedom and I'm okay with like failing, right. You know apa jadi dari situ kayak you know aku ngerasa kayak oh ya coba jalan sendiri lagi deh gitu kan.

14:04 Victor: Ini berarti founding sendiri atau sama co-founding?

Fron: Sekarang founding sendiri. Founding sendiri, founding sendiri sekarang. So sekarang model bisnisku tuh masih sebagai agency gitu jatuhnya. So I run-

Victor: A native service.

Fron: Ya native service, yes. Aku juga that's a very good point that you brought up AI native service. I think menurutku AI juga I wanna test this thesis of like dulu tuh banyak agency yang sebelum AI they need a lot of people, right kan. Mereka butuh banyak orang, mereka butuh desainer, mereka butuh orang yang actually deploy the ads to Meta, klik. It's like an actual job those clicking.

Those ad manager itu kayak it's not actually simple interface. There's like a full time people that deploy the ads. Ada lagi yang bikin strateginya gitu kan. Tapi with AI I realize kayak oh actually I can do all of that just by myself, right. Terus kayak aku juga kayak I realize kayak klienku tuh kayak one of my client mereka use different agency to deploy, use different agency to design, use different agency to come up with a strategy gitu kan.

I realize oh pas aku dateng ke mereka I can do all that. I can do all that kayak from dari strategi sampai deployment sampai tracking. Jadi kayak the creative work also the strategic thinking and also the engineering involved into it I can do everything. Terus kayak my background in engineering helps obviously with the engineering. Terus kayak AI helps a lot with the creative process, right. So I realize oh I can just make it vertical and it's just me, right.

Terus kayak and I can charge very competitive rate, right. Tapi bedanya kayak all that money goes to me, you know. Instead of kalau agency kan dibagi-bagi lagi gitu. So I realize kayak I wanna test how far I can bawa limit AI itu ke how much I can do with it.

16:08 Victor: This is very interesting ya karena one of the motivation kenapa gua mau bikin startup-nya sekarang, no sorry the podcast sekarang. It's because I can use AI to do a lot of things kan. Dulunya kan kalau lo mau bikin podcast lo mesti ada kayak video editor gitu kan. Terus ada yang clipping, ada yang posting. Sekarang kita bisa bikin kayak pipeline to do all that at once.

Fron: Aku, ya aku percaya kayak I think AI itu it's actually gift for the term tuh di sini yang people like to call it high agency. It's a gift for high agency, high creativity people because like they realize it's like oh sebenarnya I can do much more with it, right. And I think you know there's a lot of push back with AI which I think totally fair. Tapi kayak yang harus di-highlight juga kayak AI itu kayak really multiply our ability.

And I think aku juga percaya kayak you know, I guess like tujuan hidup kita di sini tuh bantuin orang gitu kan. And AI can help us to help more people, right. And those people you know if they also think tujuan mereka bantuin orang, they're gonna help more people, right. So how is that a bad thing, right, if everybody just helps more people, you know.

17:15 Victor: Ya, ya I mean that's how money works kan. Money is just a appreciation that you have show value to others life. Kenapa Apple become one of the largest market cap is because people really enjoy their devices. Why Nvidia is now the largest is because everyone wants their GPU, right. And why wants their GPU is because it's useful. Jadi kan usefulness and helping is the most important thing, right.

AI ngambil kerjaan orang? 17:41

Victor: Jadi I think I wanna push back on you a bit though you know like about why people does not like AI. It's clear that it's taking people's job. Kan kita baru ngomong sekarang gua enggak perlu hire video editor lagi. Gua enggak perlu bikin like clipping. Jadi kayak kemarin yang kita diskusi tentang the animation, right, the motion design. Claude Opus 5.5 you said just one shot, right.

So now people are losing their jobs. So ya I wonder like a you know I don't think we have a solution for this, but what's your thoughts on this?

Fron: I think I wanna acknowledge you know AI does replace some jobs and that can be bad unfortunately, right. Tapi kayak, tapi aku mau push kayak this is not new. In a sense kind of like AI ultimately is just like technology, right? So, you know, I'm pretty sure back then kayak, you know, banyak pekerjaan yang ngirim surat.

18:43 Victor: Oh man, agriculture used to be ninety eight percent of the population. Sekarang is one percent.

Fron: One percent, right. So, I think, you know, without trivializing the impact or the real impact that people have, I think what I want to highlight is like this is technology. And then I think what's different about AI itu tuh kayak, it will I think, kepercayaanku adalah it will increase the value of each individual. Assuming dia itu emang pengen ngebantuin orang. So, and I think, you know, I believe kita semua tuh kayak ada ini lah kayak gift, either our talent atau our background, whatever that gift that we have, AI itu bisa amplify that.

So that, and then we can get more value, right? Kasar katanya kayak, you know, with AI we can make more money individually, right? As long as we understand that kita di sini itu untuk bantuin orang gitu kan. So, jadi kayak aku juga wanna, I think my view ku tuh kayak I think the nature of the job is gonna change as well. Mungkin kayak I think, makanya kayak the reason aku keluar perusahaanku gitu kayak I realize kayak I think less and less people don't need to stay in a company to be able to make money anymore.

Kayak kita bisa langsung ke source of value. Kayak contoh-contoh kasarnya, contoh simpelnya kayak aku kayak instead of kerja di perusahaan lain kayak aku langsung ke klienku aja gitu. Terus aku juga notice kayak banyak yang, you know, banyak ada temanku yang dulunya kerja di perusahaan besar gitu, tapi he realized kayak oh sebenarnya gua bisa langsung ke klien langsung aja gitu kan. Terus daripada gua cuma dapat duitnya sedikit dari percentage-nya gitu kan gua langsung ke klien and with AI kayak itu kayak rasanya tuh kayak lu punya employee sendiri gitu kan.

20:48 Victor: Ya betul.

Fron: So, I think the nature of job is gonna change. I think people need to realize kayak lu harus datang langsung ke siapa yang butuh lu gitu kan. Kalau gua sebagai marketer ya klien gua butuh marketer gitu kan. Yaudah I just come to them directly instead of joining another marketing agency.

Victor: That's right. So, one thing that, satu contoh ya yang gua sering dengar ya, itu tuh kita tuh bukan dibayar tentang task ya, bukan cuma task-nya doang, tapi outcome gitu kan. Jadi sebagai programmer kita kan task-nya itu write code ya, bikin coding doang. Tapi outcome-nya itu apa? Bukan coding-nya. People don't care about coding-nya. People mau aplikasinya. People mau websitenya. Jadi kalau kita bisa langsung jual websitenya, sama kayak ini kan.

Fron: Benar, betul.

Victor: Kan advertisement itu kan cuma task doang. Outcome-nya apa? Outcome-nya para klien mau customer mau beli.

Fron: Benar.

Victor: Jadi kalau kita bisa langsung bantuin untuk beli, right.

21:50 Fron: So, like I keep saying I have this philosophy kayak ujung-ujungnya itu apa? By the end of the day ujung-ujungnya apa yang mereka mau gitu kan.

Victor: Ya betul.

Fron: Terus kayak sebenarnya kayak selain ujungnya itu kayak a lot of things doesn't matter gitu. Kayak tadi you said kayak ujung-ujungnya yang orang mau tuh websitenya.

Victor: Ya betul.

Fron: Bukan the code. Ujung-ujungnya orang mau tuh kayak in my marketing kayak orang beli produk dia lebih banyak apa enggak gitu kan. Orang inget company dia apa enggak gitu kan. That's what they care about. Kayak caranya ke sana itu kayak it matters less and less gitu kan. Jadi kayak, jadi kita bisa langsung ke value-nya apa. Nah, dari situ resikonya lebih tinggi sure, tapi kita bisa capture the value lebih banyak juga gitu kan.

Jadi kayak earning potential-nya menurutku lebih tinggi.

Cari klien pakai AI 22:50

Victor: One more question about your company ya, but ini kan baru bikin ya only one month and now sekarang kan Mas Fron kerja sendiri. How do you get customer? Like lu gimana bisa cari customer, pakai AI juga?

Fron: Ya, so yes.

Victor: Marketing, marketing agency.

Fron: So, that's a very good point. Jadi kayak ya aku pakai AI to find my client. Aku ada lah kayak I come up with a thesis kayak oke kira-kira siapa nih yang butuh service gua gitu kan dan siapa yang bisa bayar harganya tinggi gitu kan. Terus kayak I come up with like a simple rule kayak ya udah gua cari aja company baru raise funding. We live in Asia.

Banyak banget. Terus kayak dari situ kayak just itu kayak one filter kayak baru raise funding apa enggak. Mereka ada yang raise funding this past six months atau nine months. Terus dari situ dapat tuh list gitu kan. Dapat list terus ya udah aku baca, you know, each of the company from the list. Terus aku juga yang kusuka adalah aku tuh sebelum reach out ke mereka, I already create semacam kayak proposal atau strategy document.

Jadi kan aku marketing kan, jadi I come up with like, you know, I use AI to understand their company.

24:13 Victor: Of course, ya.

Fron: I understand their company terus aku baca nih, oh this is what the AI understand gitu kan. Terus dari situ I brainstorm with AI kayak oh kira-kira karakternya apa nih yang dibutuhin gitu kan. Terus dari situ AI comes up with the character. Aku review kan, oke I don't like this, I don't like this, oh I like this one, oh I like that one gitu kan. Terus si karakter kalau karakternya udah ini bikin ads-nya gitu kan.

Eh sorry bikin strateginya kayak messaging angle-nya apa gitu kan. Kayak siapa client-nya gitu kan. Terus aku juga kayak you know really understand klienku sekarang aku ngurusin banyak kayak B2B business. Jadi klienku kayak jualan produk ke another business gitu kan. Jadi kayak aku tuh juga bantuin nyari using AI bantuin nyari kayak misalkan klienku contohnya ada dia jualan voice AI. Terus aku nyari, you know, I write kind of like script using AI.

Oke find companies yang kira-kira kayak banyak call center gitu kan. Yang banyak call center di situ kayak the AI you know scrape LinkedIn kalau macam terus kayak give me like a list. Terus dari situ aku kayak come up with a strategy. Oh oke sebut saja nama company-nya I don't know Suara AI you know. Kayak dari you know Suara AI kayak, oh kayaknya Suara AI itu harus target company in a bank ini, harus target company insurance ini.

Soalnya mereka banyak call center needs-nya gitu kan. Terus dari situ aku create specific ads targeting those companies, right. And all of that I use AI with me in the loop kind of reviewing, right.

25:49 Victor: Kenapa harus pakai karakter, Mas?

Fron: Ya, so I think karakter itu kayak tadi it gives soul to the ads. Jadi kayak lebih warm aja sih kayak jatuhnya. Kayak I think that's also my thesis. Apa kayak it doesn't have to be. Some of the ads that I created enggak ada karakter cuma teks doang atau kayak just like simple image gitu kan. Tapi kayak yang aku found kayak just in terms of like you know people click or not yang ads yang memakai karakter tuh lebih di-click.

Kayak terms-nya tuh click through rate.

Victor: Ya CTR.

Fron: Ya CTR. Berapa orang yang ngelihat, berapa orang yang klik gitu kan. Terus rasionya itu the CTR gitu kan. CTR-nya lebih tinggi kalau ada karakternya. I think it kind of makes sense. Soalnya thesis ku kayak emang just dari kualitatif aku ngelihat. Itu kayak kalau ada karakter tuh merasa kayak ada emosinya, ada apa, jadi kayak lebih gampang keinget gitu kan.

Victor: I don't know ya. I, kita kan di dunia tech nih kan. Terus gua lihat di B2B companies or even B2C jarang yang pakai karakter. Kan kayak Apple, Nvidia enggak ada karakter. Databricks, Snowflake kan mereka enggak pakai karakter. Terus gimana itu, Mas?

27:04 Fron: So I think kayak I'm not saying mereka itu salah gitu. I'm just saying my approach mungkin ada value-nya gitu kan. Terus kayak maskot is not like a new thing.

Victor: No, it's ya. Kayak Ferrari ada the horse gitu kan. Marlboro ada Marlboro Man.

Fron: Michelin kan.

Victor: Ya Michelin.

Fron: So it's not a new thing gitu kan. So I'm pretty sure there's a lot of ways buat marketing yang efisien. My thesis is just happens to be from what I run from my clients is that having karakter actually helps building a brand juga gitu kan. Terus juga sekarang kayak and it's juga sebenarnya juga ngelawan this AI slop gitu. Soalnya kalau kita just from AI make an ad gitu kan.

Orangnya random lah apa gitu kan. Kontennya tuh kayak enggak ada arah, enggak ada konsistensi. So I think dengan karakter gitu kayak it's actually reducing the slop kan. Soalnya I think AI-

AI slop itu apa? Soal selera 28:10

Victor: Coba jelasin, Mas, slop itu apa? Jelasin untuk nenek. Nenek kan enggak ngerti slop itu apa.

Fron: That's a very good question. Actually now that I think about aku enggak tahu slop tuh bahasa Indonesianya apa. Tapi kayak slop tuh kayak apa ya mungkin the closest thing di kepalaku tuh kayak konten receh gitu kan.

Victor: Begitu lihat mau tahu oh ini AI gitu ya.

Fron: Begitu lihat ini AI gitu kan. Begitu lihat kayak oh ini orang pasti promenya kayak sekali doang gitu kan. Kayak terus kayak apa enggak ada kayak macam kayak fine tuning gitu kan. Kayak apa kan aku ngelihat prompting itu kadang aku kalau lagi nge-prompt aku ngebayangin aku kayak lagi bikin patung kayak mahat gitu kan. Kayak really kind of like oh ya pay attention to this, pay attention to that gitu kan.

Apa kayak and I think yang aku juga sadar itu sebenarnya I think yang orang push back itu the lack of intent from the creation gitu kan. Kan dulu kalau kita, what I mean by that kayak orang bisa appreciate animasi bagus atau gambar bagus. Soalnya kita tahu ada manusia di belakang itu gitu kan. Yang emang pengen bikin, yang pengen bikin animasinya emang look like that.

Yang pengen bikin karakternya kayak behave like that gitu kan. Kayak there is like that, that's what I meant by heart and soul. Jadi kayak ada orang yang actually intentionally kayak you know kayak kalau di animasi Pixar kayak rambutnya tuh emang bergelombang kayak gitu, gitu kan.

29:47 Victor: They pay attention to it.

Fron: They pay attention to it. I'm pretty sure there's like a full time designer that actually kind of like you know really spend hours just like making sure kayak gelombang rambutnya kayak gini, glow-nya kayak gimana gitu kan. Kayak it's labor of love gitu kan. Terus kayak AI slop itu enggak ada love-nya. Kayak they just prompt it terus kayak oh ya this is enough, send it, send it gitu kan, you know.

So I think slop tuh kayak di kepalaku yang membedakan slop atau enggak itu intent di belakang orangnya. That's my personal opinion.

Victor: I have another word ya, taste. Taste. That's a good, taste kan. Bahasa Indonesianya apa ini, taste?

Fron: Ya taste I guess like selera.

Victor: Selera, ya seleranya kan. Jadi kan begitu kita, ini kemarin gua baru ke Indo Juni, Juli kan. Terus gua ke kampung-kampung ya kayak Brebes dan segala macam tuh. Banyak tuh warung-warung bakso pakai ChatGPT advertisement. Begitu lo lihat, ah ini pasti ChatGPT, ini pasti slop karena they all look the same. There's no human taste in there, enggak ada seleranya. Mereka kayak generate satu langsung pom gitu kan. Ini Mas Fron perusahaan Mimetic gimana supaya menghindari itu?

31:01 Fron: Ya, so I think right now since sekarang aku masih sendiri kan, tapi kayak I am firm kayak I think the anti-slop apa, I'm still gonna keep the anti-slop culture, you know, even my company grow to be big gitu kan. Kayak I think simple thing kayak just have kayak tadi you said I think taste is very important, right? Taste di kepalaku tuh kayak knowing yang bagus kayak gimana, yang jelek kayak gimana.

Just have like standard, right? Just have good standard, right? And I think apa standar itu kayak menurutku kayak you just have to put in the hours ujung-ujungnya. I just work on it a lot gitu kan. And then by the end of the day you will know what is good, what is bad. Jadi kayak I value kind of like, you know, experience-nya juga kayak simple things kayak never be happy with the first result.

Kayak waktu prompting something kayak, you know, kayak jangan langsung, oh ya this is good enough. Kayak really-

32:05 Victor: Coba sepuluh biji dulu ya.

Fron: Iya coba sepuluh biji atau kayak aku juga sebenarnya my approach personally kadang sebelum I write the prompt, I think about it first. I think about it. Kayak oh kira-kira karakternya gini nih, oh kira-kira strateginya roughly apa kita coba A, B, C dulu gitu kan. Jadi kayak have like an intent, right? Jangan terlalu too much give it up to the AI. I think that's makanya that's what I meant by orang di belakang AI-nya.

Kayak lu tuh sebelum typing the prompt, do you actually think about it, right? Atau kayak just like make me an image, you know. Kayak do you really think about it, right? So I think that thinking actually I realize produces better results too, right? So I think ya ujung-ujungnya kayak, ujung-ujungnya have an intent. Kayak AI is just another technology, right? Kayak apa, kayak orang yang dulu make Photoshop to manually edit, kayak they have an intent.

They have like clear intent what to do. And AI I think will be better if you just have like a clear intent. You know, that's how you introduce taste.

3 tips biar jago pakai AI 33:13

Victor: Ya, ya. Nih Mas, jadi kan Mas Fron now solo founder di Silicon Valley working with AI every day. Coba kalau pikirin top three tips and trick ya untuk teman-teman yang wants to do better with AI. Apa ini tips and tricks-nya?

Fron: So that's a very good question. Maybe I'm leaking like my alpha here, which is fine. You know, that's why you should watch his podcast to learn this alpha. Jadi kayak I realize it's like AI itu in their nature they're just mimicking something, right? Kayak the way they train-

Victor: Yes, absolutely.

Fron: ...is mimicking something, right? So I realize to create kayak good quality let's say images or good quality video have like a good reference.

Victor: Yes, reference very important.

Fron: Good reference. Good reference gets you a long way, right? Kind of like spend a lot of time in Pinterest. So I work in images, right? Spend a lot of time in Pinterest. Kayak really oh ini mood board-nya bagus nih, oh ini mood board-nya jelek nih gitu kan.

34:17 Victor: You know what's funny? I'm actually joining Pinterest soon.

Fron: Oh, no way. Okay. Yeah, yeah, yeah. There you go. There you go. It's another, you know, I love Pinterest. I think if you're in a creative field, Pinterest it's a godsend gift, really.

Victor: Why? Why Pinterest? Because like images are everywhere.

Fron: Yeah. So I think Pinterest I like their search feature, right? I can search like very abstract thing and they just have results that is-

Victor: Wow.

Fron: ...somewhat relevant. Mungkin, mungkin aku cocok kalau [kurang jelas], oh this is relevant-nya gitu kan. Tapi kayak it doesn't matter. Kayak, you know, this is actually good direction. Terus juga what I like about Pinterest it's like when you click like an image, it will tell you image that is similar to that, right?

Victor: Yes, yes.

Fron: Itu bagus buat going down the rabbit hole juga gitu kan.

Victor: The semantic vector graphics.

Fron: So I think good reference it's a very, very good tip, right? I think I honestly, you know, like if you wanna get better at producing something with AI, that's probably my number one tip, right?

Victor: Ya, ya I think to go deeper into this reference kan. Sekarang I also like try to build games to play with my kids, right? And sekarang pakai coding bikin games is so easy, right? So if you don't have reference, kalau enggak ada reference, kita bilangin coba bikin game ini, dia tuh kayak slop ya. Jadi dia produce random stuff. Tapi kalau lo bilang, ini nih gua suka game kayak Stardew Valley, gua suka game kayak Hades, terus coba kasih gua beberapa option untuk game seperti ini, terus dia produce, terus kita bilang, "Oh gua mau kayak gini," gitu ya.

Terus begitu ada reference, they make a better result.

35:55 Fron: Yes, yes, yes. So I think reference is one. Another tip I think this is maybe more for like technical-minded people. Soalnya tadi aku bilang my background itu sebenarnya engineering. Jadi kayak by my training I think like systematically just like any engineer is required to be, right? So aku sadar kayak, you know, it's actually a good idea to have like almost like a system when you're working with AI.

What I mean by that it's like contoh kayak tadi aku pengen bikin misal gambar ads gitu kan. Aku tuh ada satu workflow yang tugasnya itu bikin ad copy-nya, kayak bilang apa gitu kan. Ada satu yang bikin visual concept-nya, ya kan. Terus dari konsepnya dikirim, dari konsep baru generate image gitu kan. Jadi kayak having itu in the back of your head, kayak strukturnya kayak gitu, gitu kan.

Victor: Ini kayak di zaman dulu kan kita ada banyak spesialis ya.

36:55 Fron: Benar.

Victor: Satu orang artis tukang gambar, satu orang copywriter tukang writing. Terus kalau ditranslasi ke AI, itu bukan satu AI bikin semua, tapi setiap AI bikin sendiri baru nanti digabungin.

Fron: Benar, benar, benar, benar. So I think, iya kayak analoginya kayak ya bayangin aja lo ada employee yang fokus copywriting, employee fokus bikin kayak conceptual art, ada fokus yang ngegambarnya gitu kan. So I think menurutku ini useful soalnya kayak sometimes the things that the AI produces actually bad. Actually not sometimes, a lot of times, right? This is talking about slop again, right? Tapi kalau misalkan sistemnya, kalau lo ada sistem, you can kira-kira menerka nih, oh ini salahnya di mana kira-kira.

Is it in the copywriting or is it in the visual concept writing, right? Terus kayak dari situ tuh lo bisa nge-debug gitu, kayak oh ini kira-kira if I improve this thing, apa yang bakal terjadi gitu kan. Jadi kayak you can reason with it. Kan soalnya kalau if you just treat AI just like a black box, it's very hard to improve things, right?

Soalnya kayak I guess better prompt ngebantu, tapi there's a limit to that, right? So I think it's like kayak apa, kayak a lot of things itu nggak bisa di one shot sebenarnya.

38:19 Victor: Iya betul, nggak bisa.

Fron: Nggak bisa di one shot. You still need to ngebagi-bagi gitu kan.

Victor: You can one shot, tapi hasilnya nggak bagus.

Fron: Benar, you can one shot and then produces a slop.

Victor: Iya betul, betul.

Fron: Sekarang aku kayak my system tuh kayak ada yang fokusnya itu kayak finding screenshots from my client's website, terus taruh di, taruh screenshot tersebut di ads-nya gitu kan. Biar kalau misalkan klien ku kayak B2B software, kayak dashboard-nya gambarnya gimana, put it in the ads gitu kan. There's a specialized, you know, AI workflow gitu kan. Dan di kepala aku tuh kayak, oh ya there is kayak ada gambaran sistemnya tuh gimana.

Ujung-ujungnya semua sistem. I believe everything is a system. There's good system, there's bad system. You just need to ensure that the system is doing good.

Victor: Jadi pertama kan reference, harus reference-nya bagus. Dua bikin sistem dan decomposition.

Fron: Decomposition.

Victor: Jadi bukan cuman satu ChatGPT window, terus lo bilang semua dibantai di sono, lo mesti dibagi-bagi. Oke, another one?

39:22 Fron: I think the last one is like, yeah I think kayak tadi aku udah bilang, have an intent.

Victor: Intent.

Fron: Have an intent, right? Kind of like really kind of like think about it, right? Kind of like, you know, I think kayak philosophically I really hate the AI slop. I just, you know, like beating that dead horse here. Tapi kayak I think, you know, like soalnya I'm very pro AI in a sense. Tapi I understand the pushback, right? And I think a lot of the pushback is because people just don't have any intent, right?

They just kind of produces kind of like, you know, trash content.

Victor: They don't have taste.

Fron: They don't have taste. They just kind of like, you know, quality matters. It's not just quantity, right? With AI, it is possible to produce high quality at high quantity, right? Tapi the quality part, intent still matters, right? So just like, you know, kayak mindset ku kayak sekarang tuh kayak be a good director. Kayak I think of myself as like a movie director or like a creative director gitu kan.

Kayak I take that job very very seriously gitu kan. I think, you know, I'd like to introduce heart and soul to the things that I created gitu kan.

40:36 Victor: Ada tips gimana untuk develop intent or taste gitu nggak, Mas? Soalnya kan Mas Fron is not a creative person quote unquote karena kan kita kan engineer ya. Engineer lebih thinking about problem solving, about structure. Tapi ini creative advertisement, animation, graphics itu kan beda. Gimana untuk mentraining diri sendiri untuk develop the taste and the intent?

Fron: Ya, so I think kayak knowing kayak the terms in whatever field that you're in. Kind of like that can be helpful. Jadi your prompt itu bisa lebih precise jadinya. Kayak contoh kayak, you know, if you're talking about web design, right, knowing kayak ada H1, ada H2, ada paragraph gitu kan. Kayak and then when you kind of like prompting, you can like hit that specific thing gitu kan.

Atau kayak kalau di, ya di ads itu ada headline, ada subtext, ada the scene gitu kan.

41:44 Victor: CTA.

Fron: Iya CTA, right, exactly. Kayak knowing that terms gets you like a long way. So that's a long way to say kayak knowing the technicals of the things, itu masih valuable.

Victor: So you have to study.

Fron: You have to study. You have to study kayak misalkan sekarang, you know, I'm trying to be creative, so I need to study kayak orang di advertising tuh terms-nya apa aja gitu kan. Orang di graphic design terms-nya apa aja.

Victor: Do you study people like Don Norman, like the affordances and the other?

Fron: So I study, sekarang aku lagi study, I don't know how to pronounce his name, Ogilvy. He's like a British, he's a legendary advertisement, right. Terus kayak, so dia banyak kayak talks a lot about how to, kayak dia tahun kayak his heyday tuh kayak sixties to eighties gitu kan. Kayak zaman dulu kayak, you know, iklan-iklannya tuh kayak Pepsodent zaman dulu, Dove, you know, kayak products like that gitu kan.

Terus kayak dulu apa, kayak it's like an, if you've watched Mad Men, if you know about the show Mad Men ya kan. Kind of like, so I'm pretty sure Mad Men is based on Ogilvy.

42:57 Victor: Oh really?

Fron: Yeah, I think, I think, I think, yeah.

Victor: Very dramatized.

Fron: Yeah, very dramatized, very dramatized, very dramatized. Tapi kayak apa, ya kayak just, you know, be, try to be like... Yang aku suka dari AI itu sekarang kita tuh bisa get to the near expert level pretty quickly, right?

Bottleneck kita tuh sekarang cuma curiosity, right?

So as long as keep asking the right question, AI will guide you to the right information.

So just like have that curiosity, have like really-

Pursuit of like excellence, right?

So and then you, I learned about Ogilvy, right? Kind of like, I didn't know about advertising industry. I learned about Ogilvy from like, from AI. Oh, yeah, I just asked-

You know, who should I learn from, right? Those guys, Ogilvy is one of them. Okay, tell me more about this guy, right? You know, like, is there any good books? You know, what are his core ideas? You know, like...

And I think from there, you just kind of like continue to-

43:57 Victor: Ya benar, pakai AI untuk learning is so easy now.

Fron: It's like, AI to me, I'm very pro AI. So like I realized in my personal life, it's just, it just satisfy my curiosity very, very well.

Victor: Ya. Sekarang kayak gua pas nyetir, gua seringkali pakai ChatGPT voice mode kan. Sambil nyetir voice mode. Terus ngomongin aja tentang like technical problems or the new stuff, right, or just brainstorm.

Terus habis itu kan ada record-nya and then we can go back to it. I think, yeah, AI is a game changer.

Fron: Yeah, AI is a game changer. And I think, yeah, I think as long as you're doing it for the right reasons-

Kind of satisfy curiosity, helping people, right, learning, you know, so.

Lahir di Houston, loncat SMA ke college 44:38

Victor: Let's change gear, ya. Jadi Mas Fron lahirnya tuh bukan di Indo, ya.

Fron: Bukan.

Victor: Jadi gimana nih?

Fron: Ya, so jadi dulu tuh my parents mereka kuliah di Texas.

Victor: Texas.

Fron: Di Texas, di Houston specifically.

Victor: Panas banget tuh.

Fron: Yes, ya panas. Ya, that's right, that's right.

Victor: Kayak Jakarta.

Fron: Yes, kayak Jakarta, tapi lebih kering sih ya dibanding Jakarta. Tapi panasnya sebenarnya panasan Texas daripada Jakarta, tapi lebih kering. Jadi in a way it's somewhat better.

Ya dulu orang tua was undergrad and grad school di Texas and then aku happen to be born in Houston, Texas. Terus jadi sebenarnya numpang lahir aja sih. I don't even remember anything. Kayak aku umur dua tahun balik ke Indonesia. So I, aku grow up besar emang di Indonesia.

Victor: Di mana?

Fron: Aku di Jakarta, di Jakarta Selatan.

Victor: Jaksel.

Fron: Jaksel. So nama daerahku tuh Ulujami, Ciledug. So there's a lot of orang Betawi di sana actually. Aku bukan orang Betawi, tapi neighborku banyak orang Betawi actually. Ya, ya, ya. So they're great. Nama daerahnya Ulujami. So I grew up, you know, I mean dulu waktu, you know, kecil pernah pindah-pindah, tapi kayak for the most part di Ulujami gitu kan. Rumah kakekku di situ.

46:00 Victor: Sampai umur berapa di Indo sebelum balik lagi ke sini?

Fron: Ya, di Indo itu aku sampai umur enam belas.

Victor: Wah, lumayan lama, ya.

Fron: Ya, lumayan lama. Jadi dari umur dua tahun sampai enam belas, you know, so fourteen years.

So that's why I think I definitely grew up di Indonesia. So like apa, ya, although I was born in the US and di Amerika udah lama gitu kan, we can talk more about that. I'm, you know, being an Indonesian is still a huge part of my identity.

Victor: Kenapa umur enam belas? Enam belas itu kan kayak setengahnya SMA.

Fron: Ya, ya, ya. So that's a very good observation. Jadi aku sebenarnya, apa, praise God, jadi aku kayak, jadi aku SMA tuh dulu di SMA Delapan, SMA Negeri 8 Jakarta, tapi programku program internasional. So jadi kita kurikulumnya itu, it's like a British based curriculum. If you heard about IGCSE or like, you know, just British curriculum gitu kan.

Victor: SMA Negeri ada British curriculum-nya? Wow.

Fron: Ya, so SMA Negeri ada internasional programnya. I got into that program and then it was British curriculum. Terus sebenarnya kelas dua SMA menaiki kelas tiga, mamaku nanya, "Kamu mau lanjut kelas tiga SMA di Amerika aja nggak?" gitu kan. Terus aku kayak, "Oh ya, sure, why not?" gitu kan. Kayak aku mikir gitu kan. Tapi terus aku being, you know, I always try to like just take action right away gitu.

I research kan, oh, yeah. Oke, how, what is the best high school in America? So I Google segala macam gitu kan. Terus kayak I found ternyata there is like a path. Kayak the test yang I took in this British curriculum, itu tuh you can convert it to like equivalent of high school diploma. Di sini tuh Amerika ada namanya GED.

47:50 Victor: Ya, GED.

Fron: GED. So it's like, so it's not really, it's equivalent of high school diploma. It's not high school diploma, but equivalent. Kayak di Indonesia mungkin kayak paket C gitu kayak, you know, kayak-

Victor: Kejar paket, ya.

Fron: Ya, ya, ya. So there is like an equivalent of it. Terus turns out I already took the test and I passed, you know. Kayak tesnya tuh kayak math, fisika, biologi, it's like various kind of subject. I can convert that to high school diploma gitu kan. Terus kayak, oh, oke.

Victor: Itu dari kelas dua SMA kan?

Fron: Ya, kelas dua SMA.

Victor: Belum kelas tiga.

Fron: Belum kelas tiga. Ya. Terus kayak I don't know if you know, ada nggak di Pacific Place di Jakarta, di mall namanya @america. If you heard about it. Ya, so it's like I think American Embassy kayak, you know, if you wanna learn more about America, you can go to @america. Terus kayak aku mau confirm nih gitu kan kayak, "Is this really true I can use this test as an equivalent high school diploma?" Terus kayak you know, aku you know kayak habis pulang sekolah pakai baju SMA ke @america gitu kan.

Terus kayak just like, terus kayak I just go there terus ask kayak ke orangnya kayak, "Mbak aku mau nanya tentang ini, who do I need to go to?" Terus kayak, "Oh coba tanya Mas ini." Oke sure. So I go, you know, kayak I remember I think I asked bunch of people to finally meet the right people that I need to ask. Terus aku nanya kayak, "Is this true?

Can I..." Ya udah aku nunjukin tes hasil tesku kayak, "So I have this. Can I convert this to high school diploma?" Terus kayak, "Oh ya bisa sih.," gitu kan. Terus kayak, oh oke. Terus aku juga di situ juga aku juga sempat aku dengar research namanya Community College. Jadi kayak apa, two and two put together kayak oh aku punya equivalent high school diploma, I can use that buat apply Community College di Amerika gitu kan.

So I can skip a year of high school. That's why I moved when I was sixteen. Jadi aku dari kelas dua SMA langsung ke college, Community College.

49:47 Victor: Di mana itu?

Fron: Dulu di Houston.

Victor: Oh Houston lagi.

Fron: Namanya Houston Community College.

Dikirim ke Texas, masuk UT Austin 49:52

Victor: Jadi pas itu Mas itu orang tua ikut ke sini lagi atau di Indo lagi?

Fron: Enggak. Jadi tadinya aku pengen banget ke California because it's California. I heard great stuff about California.

Victor: Of course, we live here now.

Fron: Ya, ya, ya. Terus aku bilang kayak, "Oh ya, Ma, sebenarnya bisa nih aku langsung skip senior year, gak perlu SMA lagi, langsung ke Community College," gitu. Aku nanya-nanya, this is like a good Community College. Terus sistemnya kayak, jadi for viewers that don't know, jadi di sini tuh ada option you go to Community College, habis itu transfer ke university. And then you finish your S1-nya di university. Jadi mulai di Community College, finish di university.

Terus I told my mom that kayak, "Oh ya, this is what I found," gitu ya. I can just go straight to Community College instead of finishing high school gitu kan. Terus tapi mamaku, "Oh ya oke, tapi kamu ke Texas ya," gitu kan. Terus kayak, "Oh why Texas?" gitu kan.

Victor: Iya kenapa?

Fron: Terus kayak soalnya I guess mamaku kayak, I guess looking back it was understandable. Jadi kayak mamaku dulu like I said S1, S2-nya di Texas gitu kan. Jadi banyak teman di sana. Jadi mamaku, me being still sixteen, mamaku lebih comfortable untuk ngirim aku ke rumah one of her friends.

51:09 Victor: Biar enggak terlalu liar ya.

Fron: Biar enggak terlalu liar atau kayak, exactly, biar enggak terlalu liar. Terus juga kayak lebih terpercaya aja gitu kan.

Victor: Ya betul, betul. Understandable lah.

Fron: Ya understandable. So oke, Texas. Tapi what's funny emang dari dulu aku tuh kayak I'm not sure if my parents is doing this propaganda intentionally or not. Kayak dari kecil tuh kayak mamaku sering bilang kayak, "Iya mama tuh dulu kuliah di Texas, tapi teman-teman mama yang pintar-pintar tuh pada ke University of Texas," gitu kayak they're all very smart gitu. Aku dari kecil kayak I heard about University of Texas dari kecil gitu kan.

Terus kadang kan kalau orang ditanya kayak, "What's your dream college?" People say, "Oh Harvard, Stanford, whatever." I was like, "No, I want to go to University of Texas," you know.

Victor: Propaganda.

Fron: The propaganda works, man. I don't know. Kayak you know, terus kayak ya udah terus kayak mamaku ngasih, "Ya udah kamu ke Texas." Terus aku, "Oke, if I'm going to Texas then my goal is to go to University of Texas because apparently that's where the smart people in Texas go," gitu kan. Terus kayak ya udah terus kayak I ended up apa tinggal sama teman, kayak teman keluarga lah.

You know I call him uncle kayak you know we call everybody om and tante gitu kan. Tapi kayak I feel like they are the closest, buatku closest om yang bukan om kandung gitu.

52:29 Victor: Closest not blood related.

Fron: Closest not blood related. That's right. So I stay with them, apa they were very kind. Kayak honestly kayak aku bisa tinggal di situ, aku bayar rent, tapi mereka ngasih makanan juga it is part of the deal. Terus kayak some part of me thinks I think they're losing money by taking me in gitu kan. So they were so kind. Mereka baik banget. Ya terus kayak aku di, ya udah di-introduce to American culture through their household lah gitu kan.

So I moved ke Houston 2012 when I was sixteen. Ya di situ kayak I was in Houston for a year and then terus kayak transfer ke UT Austin. So praise God I got into the, my target school.

Victor: The propaganda works.

Fron: Ya propaganda works. It's like, you know, I, you know, I'm still, you know, a Longhorn, you know, like I'm a big Longhorn fan.

Victor: Jadi University of Texas, Austin itu mirip sistemnya sama UC ya, UC Berkeley, UCSD gitu ya.

Fron: Ya, ya. So jadi kayak, ya so University of California and University of Texas is like very similar.

53:30 Victor: Mungkin untuk the reader in, oh the watcher in Indo, kayaknya Indo gak ada sistem kayak gini deh. Ini kayak tiap negara bagian ada sekolah negeri yang bagus banget. Bagus banget. Jadi UC Berkeley kan bagus banget, tapi itu punya negeri dan juga ada yang punya swasta.

Fron: I think the term is flagship.

Victor: Flagship, iya. Kayak pentolan, kayak who's the top dog in each system, right?

Fron: Ya, ya, ya. Jadi Texas tuh top dog-nya University of Texas. Like you said UC is UC Berkeley. Not to say the other school are not good. Kayak banyak, there's UT San Antonio, UTP, di Texas juga ada Texas A&M. Our good friends Randy actually went there.

Victor: Oh Randy, really?

Fron: Randy went to Texas A&M. So, tapi kayak flagship just means it's the biggest and mereka juga tend to get the most funding.

Victor: Ya, ya, ya. So before we go into the university and what you study, ini coba give some reflection ya dari Indo itu kenapa lu bisa dari kelas dua SMA bisa pass tes yang buat kelas tiga. Kok bisa?

54:40 Fron: Ya, so I think sebenarnya menurutku kayak itu lucky aja sih. Kayak apa jadi what I mean by lucky it's like sebenarnya mamaku yang nyuruh, "Yaudah kamu ke SMAN 8." So SMAN 8 happens to be like, you know, one of the more competitive ones lah di... Terus mamaku bilang, "Kamu SMAN 8 ke international program." It's okay. I was like, "Okay, sure." You know, I got into the international program gitu kan, with no intention of like using that for anything.

I just wanna go to good schools, you know. Terus kayak yaudah I was in a international curriculum. It's like apparently this British curriculum, this IGCSE, it's like a Cambridge punya sistem gitu kan. So it's very like, it's very international, apa well-known. It's like a standard lah gitu kan. I think Singapore, you know, adopt this curriculum, you know, obviously the UK, the Oxbridge, right?

I think, you know, a lot of some schools in America even kind of like adopt this curriculum gitu kan. Jadi kayak I would say it's just like, you know, credit to my mom, really, like telling me to, "Okay, you should do this," gitu kan. Terus kayak from there kayak apa, ya dari situ kayak, you know, learn more about the curriculum. Jadi sistemnya itu kayak sebenarnya aku ngambil ujiannya itu kelas 10 which is like what-

56:01 Victor: Kelas 10.

Fron: Sophomore here, like the tenth year. Ya, kelas, ya, ya. Kan kalau di Amerika kan high school itu 9, 10, 11, 12 kan.

Victor: Ya betul.

Fron: Kalau di Indonesia 10, 11, 12. Jadi 1 SMA. So I took this exam actually 1 SMA. So I'm very surprised that, you know, there is a path to convert that to high school diploma. Honestly, this is like another alpha. Like if you're like a high school student trying to kind of skip. So potentially you can skip two years. So aku ngambilnya kelas 1 SMA. Jadi sebenarnya I learned teman SMA-ku ada yang kelas 2 tiba-tiba dia pindah ke Amerika.

Kayak turns out this is what she did, which is, you know, she just didn't share to her friends, you know. So this is another alpha. It's like kamu bisa ngambil IGCSE, convert that to high school diploma and go to community college. So yeah, honestly kayak it's all just, you know.

Victor: Kemarin pas lo ngetes itu kerasa susah apa enggak?

Fron: So the te-

Victor: Atau preparation-nya misalnya, belajarnya.

57:02 Fron: The test itself is just like, kayak it's just another test, right? You know, it's just like you just need to, it's like good thing it's part of the curriculum. So this is what we do day to day-

Victor: I see.

Fron: Learning in class. So this is literally just like, you know, you learn about the cur-- apa you learn about... Basically you prepare the test the entire year.

By learning math, physics, biology, whatever, you know. So jadi kayak emang sekolah curriculum ini designed untuk, to take this test.

Victor: I see, I see.

Fron: So tapi I'm not sure if you can take the test yourself without part of like a school. Maybe. So maybe that's like another option.

Victor: Jadi just randomly take some GED gitu?

Fron: Ya, soalnya kayak ada temenku yang dia, he's not part of the international curriculum, tapi he took the test.

So I think it's possible. So maybe you just need to find some agency that where you can, that maybe accredited agency where you can like, you know, take the test and then... So karena mereka butuh ini kan kayak trust on like lu tuh ngambil tesnya enggak curang segala macam. So you need to get accredited somehow.

58:09 Victor: Ya, the proctoring.

Fron: Ada proctor-nya segala macam, you know.

Kumon dan kebiasaan matematika 58:14

Fron: So-

Victor: Ada itu enggak pas kecil gitu ngeles-ngeles gitu penting enggak ngerasanya?

Fron: Oh man, ya. So I wanna give shout out to, again, first of all, give shout out to my mom who put me into this thing.

Victor: Tante, thank you, Tante.

Fron: Yes. Ya. Ma, I love you. No, I think, I don't know if you know, but ada Kumon.

Victor: Ya, Kumon very famous. Yes, yes.

Fron: I think Kumon really, you know, in the moment I hate it. I hate it going through it. But tapi sekarang ngelihat ke belakang kayak, oh man, I think, you know, Kumon definitely, you know, set me a life, set me up for life for good, really. Kind of like, soalnya kayak I'm a big believer kayak math education is actually very important.

Kind of like satu dari, it trains you to see the world in a rigorous way.

Right? You can, you know, see the world break apart, see how each thing kind of related to one another.

Victor: Systematic.

Fron: Systematic, right. Kayak kan, you know, you learn that in kayak algebra, persamaan gitu kan. There's like, you know, this equals this, right, therefore it follows that this is the case gitu kan.

59:24 Victor: There's a really good book called Thinking in Systems, right?

Fron: Thinking in Systems, yes. I love that book.

Victor: Ya, it's similar to this.

Fron: I read that book. It's very, very good, right. So I think, ya my mom dulu naruh aku di Kumon itu kayak it's again, in the moment I hated it, tapi looking back kayak it set me up. Soalnya kan Kumon juga yang aku suka dari Kumon sistemnya adalah even when you're, even kalau kamu masih muda pun, kayak I know there is like when I was waktu di Kumon itu ada kayak anak masih kayak bocah kayak 4 SD, but they learn calculus.

Victor: Wow! Crazy.

Fron: It's crazy, right? So soalnya kayak sistem mereka emang kayak if they can move that fast-

You know, why not?

Victor: It's individualized, right?

Fron: Individualized. Very individualized. Terus juga sistemnya kayak apa, ya jadi setiap anak itu punya programnya sendiri. Terus dia ke Kumon itu ke tempat lesnya kalau misalkan kalau ada mau pertanyaan, misalnya nanya gitu kan, atau butuh more guidance gitu kan. Tapi each kids itu kayak punya individualized program gitu kan. And there's like, you know, five to six teachers where you can go to, to ask questions gitu kan.

So jadi kayak I realize kayak apa sebenarnya tuh kayak if you're actually really like, you know- apa really like the really like math, you can go, you know, as far as you can even when you still young gitu kan.

1:00:46 Victor: Itu Kumon cuma math doang atau yang lain juga?

Fron: So aku dulu math doang. Aku dulu math doang, you know, tapi kayak dulu aku kayak like I'm not that extreme kelas empat SD belajar calculus, tapi kayak but I was like let's say like two three years ahead of my designated grade. Jadi aku mulai tuh kelas empat. Kelas empat terus kayak by kelas enam tuh kayak aku udah belajar kayak pelajaran kelas sembilan gitu which is kayak, you know, like advanced algebra gitu kan.

Kayak menurutku itu bagus actually satu it builds confidence. Jadi kayak this kayak apa unfortunately banyak orang yang mikir I'm not a math person, right? I personally really don't like that idea. Soalnya kayak, dan itu menurutku dari Kumon. That because, so Kumon menurutku filosofinya bagus. Jadi kayak filosofi Kumon itu kayak you just have to make small progress every day, right? Kayak semakin tinggi levelnya, kayak dulu misalkan kayak waktu baru belajar tambah-tambahan kurang-kurangan kayak PR-nya tebal gitu kan.

Soalnya isinya cuma tambah-tambahan kurang-kurangan. Tapi lama-lama once you learn advanced calculus, sometimes PR-nya itu cuma satu lembar. Kayak cuma satu problem, tapi they emphasize to do it every day, right? What I mean by that it's like or what I'm trying to say, saying all that is like you can build a good math habit, right? Kayak you can, apa, kayak if you just do small problems every day, kayak you will be math person.

Kayak there's no way around it, right? Terus jadi kayak I don't think I'm particularly smart in a sense, apa, kayak just like have like, you know, apa my pure software have like very high IQ, tapi kayak I believe in a good habits, right, good habits can-

1:02:28 Victor: Good process.

Fron: Good process, right, good process thinking. Kayak I know someone who's like, kayak aku banyak teman yang dia waktu di SMAN 8 mereka kayak olimpiade math gitu kan. Kayak they're like genuinely intelligent, kayak there's, kayak we just have different software. Tapi kayak I could kind of like, you know, be in the same space as them because I just have good habit, right? So not because like I'm like in particular in a smart way.

Jadi kayak I believe and, you know, aku bersyukur banget Kumon really instill that point of view, right? So-

Victor: So Kumon if you need a spokesperson.

Fron: That's right. You know, if you need a spokesperson, more than happy, you know. More than happy, actually, you know. So to, so yeah, I think ya itu menurutku membantu banget sih growing up. Jadi kayak builds confidence in my math ability. I always see myself as a math person. Terus kayak, tapi ujung-ujungnya math person just means you just be able to seeing things in systematic way, right?

So that you can reason about things gitu kan. Kayak lu bisa ngeliat kayak, you know, like I see marketing as like a, you know, system, you know. It's like a system, you know. There's ad copy, there's target audience, right? It's all just like different entity that I need see, to see relationship gitu kan. So, and that takes you like far, right? And honestly, you know, it doesn't take a lot of like IQ to be able-- Once you see it, the conclusion kadang tuh kayak, oh simpel ya kalau lu tahu kayak, oh ya, kayak tadi kayak gua mikir kayak, cara nyari klien gimana?

Kayak ya gua mikir from first principle aja. Oke, I wanna be able to charge my service very high. That means I need to find klien yang cukup banyak duitnya gitu kan. How do I do that? Oh, funding, you know. So just like simple, as simple as that, right? Kind of like, and I think I give that to my, to Ku-I give that credit to Kumon, you know, like help me be able to see the world that way, you know.

Kuliah pure math, buat apa? 1:04:23

Fron: So yeah.

Victor: Terus nanti di UT Austin also study math ya?

Fron: Yeah. So I happen to study math-

Victor: Kenapa study math?

Fron: So I think in the beginning I was kind of like, you know, not sure computer science or math gitu kan. So apa I decided kind of like, I think math it's in a way it's more fundamental, right? Kind of like, apa, kayak it's just like, apa, kayak the things that I learn in math degree itu maybe it's not directly applicable to like a job, tapi kayak the thinking behind it, it applies to all kinds of job, really, right?

Terus kayak di math specifically waktu aku di UT itu kayak actually specialize in pure math. Jadi kayak-

Victor: What does it mean pure math?

Fron: So pure math it's like, jadi ada pure math, ada applied mathematics kan. Applied mathematics you use math to solve applied problems. Kayak in engineering, in physics gitu. Jadi ada real aplikasinya. Pure math it's really purely kind of like thinking about math for math sake. What I mean by that is like contohnya kayak, bahasa Indonesianya apa ya, proof, pembuktian. Kayak-

1:05:42 Victor: So people like Terence Tao-

Fron: Terence Tao, ya, ya.

Victor: Is pure math person.

Fron: Pure math, pure math person. Jadi kan sebenarnya matematik tuh, again going back to, matematik tuh sebenarnya sistem kan. Ada aturannya gitu kan. Jadi kayak, jadi math itu kayak if you wanna, you know, really take it apart, ada axiom gitu. Axiom itu kayak the ground truth gitu kan. Kayak ada-

Victor: Proven.

Fron: Ya, ada axiom kayak, you know, abis satu, dua, abis dua, tiga, selalu ada, you know, kayak superseding gitu kan. From there, kalau, dari rule itu apa yang kita bisa derive gitu kan. Yang kita bisa derive kayak, oh ada odd and even number, right? From that, what can we derive gitu kan. Kayak orang menanya kayak how many even numbers are there? Is it infinite gitu kan.

Kayak, so apa kayak mungkin pertanyaannya kayak abstrak gitu kan. And you know, like if you say itu enggak ada aplikasinya kayak I would probably agree gitu kan. Kayak the, you know, kayak, you know, why does learning about even and odd numbers, you know, is applicable. Tapi kayak my point it's like the rigor really trains me to really, you know, have this point of view of like everything sebenarnya bisa di-breakdown.

And then you can really see kayak axiom-nya apa ujungnya. Kayak, you know, Elon Musk, you know, sering bilang kayak first principle thinker gitu kan. Kayak dia selalu bilang kayak sebenarnya everything tuh apa, first principle-nya apa. Kayak apa prinsip-prinsip yang bisa kamu pakai, you know, kan. Dari situ you can reason up gitu kan. You know, so and math really, really trains me to do that, ya.

AI dan soal Navier-Stokes 1:07:15

Victor: So I want to ask, ini agak beda, ya. Baru-baru ini banyak lagi news kalau AI model ini kayak breaking or solving kayak millennium problem dan yang lain-lain gitu kan. Do you follow that?

Fron: Yes, I follow that.

Victor: Ya, jadi coba dijelasin dikit nih mungkin ke yang belum pernah dengar. Jadi why, kenapa ini important?

Fron: Ya, ya. So I think, jadi to put some context, millennium problem is just like a set of problems. I don't know who this, who's the komitenya siapa, tapi kayak let's say there's like a millennium committee, you know, like maybe, maybe hundred years ago decided kind of like, oh yeah, these are the problems yang we just think it's very important, right, and it's very, very hard gitu kan.

Victor: Dan ada duitnya juga kan?

Fron: Ya, ya ada hadiahnya.

Victor: Berapa itu?

Fron: Million dollar.

Victor: Million dollar per problem kan?

Fron: Per problem. There's like seven, I believe.

Victor: Tuh, kalau yang mau duit tinggal solving ini dapat satu juta dollar tuh.

Fron: That's right, that's right. Satu juta dollar. Ya, just solve it. Terus kayak, so this problem, apa, let's just say I don't know how they decide which problem is important or not. Tapi kayak this problem I think it's fair to say that if we actually can solve them, there is a lot of like implications from it gitu kan. Kayak, so I guess like yang kemarin OpenAI or Claude, whoever gets the credit.

It's like there's a drama there, right?

1:08:45 Victor: Yes, yes.

Fron: There's a big drama siapa yang dapat kreditnya segala macam. You know, we can talk about it. Tapi kayak yang I know is OpenAI is the one who first publish-

-the findings, right. And I think, you know, I, like, you know, like, you know, I don't wanna claim to know the details too much. Tapi kayak problem-nya mereka solve itu namanya Navier-Stokes.

Victor: Ya, Navier-Stokes.

Fron: So, Navier-Stokes ini I guess, I guess to make it, apa, easier to understand aplikasinya itu banyak di computational fluid dynamics. Kayak think of kalau kamu lagi ngedesain airplanes, lagi ngedesain F1 cars, you know, lagi ngedesain mobil gitu kan. Kayak you need to take into account kayak the flow of like waves, right.

Victor: Ini mungkin ada satu yang bakal orang nanya, why is fluid affecting planes? Karena kan airplane di udara bukan di air, bukan di water kan. Jadi-

Fron: Ya, but I think that's pretty simple to explain, right. The plane moves and it's against like a, you know, a current of wind, right. So kalau kamu pernah naik pesawat kan sometime turbulence happens gitu kan. Kayak and there's a lot of researcher trying to figure out how to predict exactly more accurately when turbulence happen or not, right. How to-- kan turbulence kayak going up and down gitu kan.

I have a good friend who's like his expertise is like making a software yang bikin pesawatnya itu kalau ada turbulence ke atas software-nya kayak-

1:10:17 Victor: Adjust ya.

Fron: Adjust gitu. Jadi kayak, jadi inside pesawat itu just selalu stabil gitu kan. So self stabilization gitu kan.

Victor: Wow.

Fron: And you need to know a lot of about fluids, right.

Victor: And this Navier-Stokes problem-

Fron: Navier-Stokes problem is relevant towards that, right. So tapi juga I heard juga kadang, apa, I heard kayak OpenAI itu solve like a very specific, apa, instance of a Navier-Stokes. So jadi Navier-Stokes can be broken up into like, you know, a bigger set of problems and OpenAI solve like one of them, right. So there's a lot of, you know, I don't wanna claim to know all the details tapi kayak, you know.

Victor: Kenapa ini susah, Mas? Jadi kan sebagai bukan orang matematika kan, gua nggak ngerti kenapa ini susah. Biasanya kan matematika kita ada aljabar atau ada kalkulus, kita tinggal masukin ke matrix, multiply, selesai. Kenapa ini susah?

Fron: That's a very good question. I think from my understanding, again, I'm not like a computational fluid dynamics people.

1:11:17 Victor: No, I just in general kan, like millennium problem dan lain-lain.

Fron: Ya, so I think jadi kayak di matematik tuh ada istilahnya, ada istilahnya tuh analytical sama computational, oke. Jadi analytical tuh kayak lu tuh bisa kayak on paper, kayak solve it gitu kan. Kayak you just do calculus gitu kan. Kayak you actually follow apa ya, whatever it is yang lu pelajarin. Atau kalkulus satu, dua, tiga gitu. You do integration by parts, whatever gitu kan.

To solve the integral lah atau to solve the differential equation. Kayak you can kayak do the algebra yourself lah gitu kan. Ada yang solve it itu using computational gitu kan. Jadi kayak, jadi apa, kalau contoh konkritnya mungkin kayak, oke, gua pengen tahu integral of this weird functions value-nya dari nol sampai sepuluh ribu dua ratus gitu. Because I care, I don't know, maybe when designing like something sepuluh ribu dua ratus is like my boundary gitu kan.

Kayak I don't really need to solve the algebra, you just give me like, you just give me kind of like the estimation gitu kan.

Terus, so my understanding is like being able to solve this Navier-Stokes will give us like better and better accuracy if not exact answer of a lot of this-

1:12:44 Victor: Fluid dynamic

Fron: ... math pro-- fluid dynamics problem gitu kan.

Victor: I see. I see.

Fron: Jadi kayak, jadi kayak... And the implication is like we'll have safer planes.

So, you know, I don't know, safer bridge, you know.

Whatever you apply the fluid, right. Yeah.

Victor: Ini- So- ... mungkin bisa jelasin soalnya gua masih enggak ngerti. Soalnya kayak someone like Terence Tao. Dia sih mungkin kayak enggak ngerti, belum pernah dengar Terence Tao is one of the yang paling terkenal lah sekarang. Yeah. Paling ngetop mathematician. Kenapa dia enggak bisa nge-solve? Kenapa OpenAI bisa solve dan dia enggak bisa?

Fron: So I think, to be my understanding, what OpenAI solve it's like-- jadi kayak contoh gampang gini. Jadi ada, you know, this equation, right. Terus kayak kalau dulu belajar aljebra gitu kan, ditanya kayak solusi persamaan ini apa gitu kan.

X bali-- X-nya berapa gitu kan.

Victor: Iya. XY.

Fron: You know, kalau X-nya dua dan empat fungsi ini nol gitu kan, you know.

So like the-- so what OpenAI did, my understanding is like they found the solution to some very complicated equations.

Navier-Stokes equations. They found like the solution tuh kayak, adalah kayak, you know, dua koma dua empat X square bla bla bla, kayak very complicated spiral gitu kan.

They just satisfy the property of the equations gitu kan.

Tapi kayak, and when you look at the solution, kayak mungkin you've seen this-

1:14:14 Victor: The spiral, yeah

Fron: ... the picture of the spaghetti, right? So that is the solution, right. You know, kind of like, whatever that means, right.

Victor: Yeah, yeah. I understand why.

Fron: Yeah, yeah. So basically kind of like, apa, so that function kind of satisfy the the equality, right.

Let's put it, you know, again, there's a lot of mathematician will roast me. Tapi kayak- ... you know, if you're really, really dumbing it down, that's kind of like you know, how it looks like. Terus kayak as you can see kind of like it's very complex gitu kan.

It's very hard for pure human to come up with that very weird functions gitu kan.

Victor: I see. I see.

Fron: Nah terus AI can come up with that because they just keep-- kayak I heard like OpenAI uses like ten thousand agents.

Victor: Yeah, ten thousand agents.

Fron: Yeah, ten thousand agents to kind of like, you know, and they spend like fifty million dollars of compute, right.

You know, so it's like long process.

And I'm pretty sure may-- or maybe not pretty sure. Tapi I'm guessing what happened it's like the AI just trying bunch of stuff-

Victor: Yeah, brute force gitu ya

Fron: ... and then bam, you know.

Like, so maybe it's not pure brute force, it's like smart brute force. But it's still like, you know, kayak something that human cannot do.

Like a single Terence Tao, no matter how smart Terence Tao is maybe could not even imagine that is the direction-

1:15:27 Victor: Tapi ten thousand Terence Tao maybe kayak-

Fron: Yeah, ten thousand Terence Tao maybe, right. Ta-tapi juga, you know, kayak these ten thousand agents they communicate real time juga kan.

So there's also like a limit on like even if there is like ten thousand Terence Tao, it will take time for one Terence Tao to talk to another Terence Tao gitu kan.

So I think that's, you know, this, that's the good way maybe to illustrate why Terence Tao cannot do it and the AI can do it.

So solusinya kayak it's kinda ugly, you know. Like, oh, wow, turns out this is-- and kita juga enggak tahu what does that mean, right.

So I think Terence Tao kemarin nge-push kayak the point of math is not just to solve the equation. The point of math is to really understand why.

Victor: Understand, yeah.

That's right.

Fron: So that we can actually apply it, right.

Victor: And now we can solve it, but we don't understand why.

Fron: Yeah. So-

Victor: Kind of stupid

Fron: ... so we can solve it. So yeah, so may-maybe like, you know, my friends who's like in the field can correct me. Tapi kayak my-- from the tweets that I see, you know, the pushback from the mathematicians is like it really defeats the spirit of math, right? Of like math is all about ultimately understanding the system, right? If we just come up with a solution, kayak you know, what does that mean?

1:16:32 Victor: Ini balik lagi kayak yang awalnya kali, enggak ya.

Fron: Ya.

Victor: Soalnya sebagai musician lo kayak ada intent mau produce musiknya begini. Lo sekarang pakai AI udah bisa bikin musik yang bagus banget.

Fron: That's right. That's right.

Victor: Sama kayak grafik juga kan.

Bikin graphics, bikin character.

Artist tuh mesti mikir, mikir alasan.

Fron: Benar, benar.

Victor: Tapi AI langsung one shot.

Mirip kayak gitu ya.

Fron: Ya, mirip kayak gitu. I think that's why I think di kepalaku kayak ujung-ujungnya kayak, you know, the goal needs to, for us humans to help another human.

And, you know, like I think, I completely get kind of like the pushback against all this AI company where they kind of like explicitly bilang kayak, "Oh, we're gonna replace you." I think, I don't think that's like a good messaging, right.

I think a better message is like, you know, is a tool to help us to help one another, right.

We are still the, let's say the main character.

You know, kind of like, you know, apa, and then without the heart and the soul of the intent or the desire untuk bantu orang gitu kayak you know, I think, you know, that could be bad, right.

So but with that, if as long as we want to help people, as long as we, there's a heart and soul, I think AI can be wonderful technologies. Yeah.

1:17:52 Victor: Well, Mas Fron, it's almost one and half hour now.

Fron: Wow! Yeah, wow, wow, wow.

Victor: Yeah. It doesn't feel that way. I know.

Fron: No, it's great. You're a great host. You know, you always come up with a great question, so.

Victor: Yeah. So we'll have to do an episode two at some point-

Fron: Yeah, wanna, wanna, yeah

Victor: ... because I don't think we should keep going.

So thank you so much. I think today we learn about Mimetic.

You know, and the top three tips, right. So jadi kan kita pertama ada reference, terus ada system-

Fron: System

Victor: ... terus ada intent or taste gitu kan.

Terus balik lagi ke Indonesia, SMAN 8.

Terus kita ada tes GED.

Fron: Yep, yep.

Victor: Jadi teman-teman yang punya anak yang masih awal bisa tuh cobain kan tes GED.

Terus kirimin ke US.

Fron: Yes, yes, yes, yeah.

Victor: Enggak usah selesaiin community college.

Fron: That's right. That's right.

Victor: Ya, ya, ya. Then mathematics. So next time we'll have to talk more about the experience di university itself.

Terus habis lulus kan-

Fron: Boleh, boleh

Victor: ... kerja di Google ya kemarin.

1:18:53 Fron: Yes, yes, yes.

Victor: Ya, terus sama kerja di startup lain.

Kita belum sentuh sama sekali nih.

Jadi we still gonna talk another time. So thank you so much, Mas.

Fron: It was a pleasure.

Victor: Yeah, yeah. This is fun.

So we'll do another one.

Fron: Yeah, cool.

Victor: Thank you.