The Everyday Miner Podcast

CellHasher founder Luke details 3,000-phone mining operation yielding $6k-$7k monthly via Varis, AI, and scraping

53 min Listen to the original Episode page
Bitcoin MiningCell Phone MiningAI InfrastructureDecentralized ComputeCryptocurrency MiningWeb ScrapingHardware Efficiency

Summary

Luke from CellHasher discusses scaling a 3,000-phone mining operation from an initial 100 devices to a commercial network. He explains that high residential power bills drove the shift to low-wattage phone mining, which offers versatility compared to ASICs. The operation generates revenue through cryptocurrency mining (Varis, Monero), AI inference for Google-funded research, and web scraping services. Luke highlights the energy efficiency of Snapdragon chips and the economic viability of distributed compute for lightweight AI agents and data collection tasks.

Episode sections

  1. 0:00 – 1:41

    Introduction and Luke's Background

    Host introduces Luke from CellHasher, who shares his history in tech, early crypto adoption via Litecoin and Bitcoin, and interest in low-power computing.

  2. 1:41 – 7:46

    Origin of Cell Phone Mining

    Luke explains starting phone mining in 2020-2021 due to high power bills, buying 100 phones on eBay, and the months of tinkering required to power them without batteries.

  3. 7:46 – 13:26

    Early Hardware and Varis Mining

    Details initial setup with Orbic Joy and AO3S phones mining Varis, noting modest daily returns but significant token appreciation that enabled scaling.

  4. 13:26 – 18:18

    Scaling and CellHasher Business Model

    Growth from 100 to 3,000 phones led to founding CellHasher, which manufactures specialized hardware; claims nearly 30,000 phones shipped and 100,000 managed.

  5. 18:18 – 22:04

    Google Partnership and AI Workloads

    Reveals Google partnership funding USC research using recycled phone compute for grading software, highlighting efficiency for small tasks vs. GPUs.

  6. 22:04 – 30:05

    Hardware Efficiency and AI Potential

    Compares phone hardware favorably to Raspberry Pis, citing Snapdragon chips with NPUs running on <5 watts, ideal for lightweight AI agents.

  7. 30:05 – 33:27

    Supply Constraints and Longevity

    Argues phone supply will remain robust compared to GPUs due to mass production and recycling, predicting technological solutions for storage/memory constraints.

  8. 33:27 – 36:45

    AI Workloads and Token Generation

    Discusses running 30B parameter models on Red Magic phones at 30-40 tokens/sec, viable for distributed agent tasks despite low latency.

  9. 36:45 – 39:13

    Web Scraping and Data Collection

    Identifies web scraping as a major use case, using SIM-equipped phones to mimic humans and evade bot detection for AI training data.

  10. 39:13 – 43:28

    Click Farming and Proxying

    Explains automated click farming for ad revenue and legal VPN proxying, distinguishing from illegal SIM farming.

  11. 43:28 – 45:37

    Operational Costs and Revenue

    Details 3,000-phone fleet costs ($400/month power) and revenue ($6k-$7k/month) from Ackerast, mining, and game servers.

  12. 45:37 – 47:54

    Midnight Network Gamified Mining

    Recounts earning $45,000 in Night tokens from 200 phones in a gamified event, warning about lockup period risks.

  13. 47:54 – 49:25

    Hardware Versatility vs ASICs

    Contrasts phone versatility across mining, AI, and tasks against ASIC specialization risks, noting phones are less risky if algorithms change.

  14. 49:25 – 52:41

    Target Audience and DIY vs Managed

    Suggests phone mining suits Linux enthusiasts; recommends managed charging stations for ease of use over DIY setups requiring technical troubleshooting.

Assets discussed

AssetKindStanceContext
LitecoinLTCcryptoneutralAt the time, it was like Litecoin and Bitcoin was like the main ones that you could buy with. So I was like, how do I get this? Like, what is Litecoin and Bitcoin?
BitcoinBTCcryptoneutralAt the time, it was like Litecoin and Bitcoin was like the main ones that you could buy with. So I was like, how do I get this? Like, what is Litecoin and Bitcoin?
EthereumETHcryptoneutraldeveloped for Litecoin as well as some things in Ethereum, like Ethereum Name Service, and a few other uh different dApps and stuff like that.
AlphabetGOOGLequityneutralwe did a deal with Google recently, which is huge. I I count that as like a a big success.
NVIDIANVDAequityneutralIt's not like all sudden NVIDIA is like we launched 10 different foundries.
MoneroXMRcryptoneutralsome people are using um phones for like Monero.
CardanoADAcryptoneutralLike a couple of years ago, um, Charles Hawkinson, the creator of Cardano, or one of them, um, there's a few, uh, he made kind of a push with this

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Luke has been in the crypto/tech space for over a decade, starting with building PCs for World of Warcraft and buying in-game items with Litecoin and Bitcoin.

Unidentified speaker (speaker_1) ·0:42

How to verify: Cross-reference Luke's stated timeline with public records of World of Warcraft's release and the early adoption period of Litecoin (2026) and Bitcoin (2026). For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Luke started phone mining around 2020-2021 due to high power bills and bought 100 phones on eBay.

Unidentified speaker (UNKNOWN) ·1:48

How to verify: Check eBay sales records for bulk phone purchases in 2026-2026 if available, or verify Luke's public YouTube channel for videos documenting this period. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Luke mined Varis for 1.5-2 years and the token appreciated 600-1200% during that time.

Unidentified speaker (UNKNOWN) ·11:48

How to verify: Check historical price charts for Varis (VARIS) on crypto data aggregators like CoinGecko or CoinMarketCap for the period 2026-2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

CellHasher has shipped nearly 30,000 phones and manages a network of around 100,000 devices.

Unidentified speaker (speaker_1) ·17:57

How to verify: Review CellHasher's official website, press releases, or LinkedIn company page for shipment statistics and network size. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Google funded a USC research project using recycled phone compute for college grading software.

Unidentified speaker (speaker_1) ·18:38

How to verify: Search USC press releases or Google.org grants database for projects involving 'recycled compute' or 'phone mining' and 'grading software'. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Modern phones have 8-core Snapdragon processors running at 2-4 GHz with NPUs, consuming less than 5 watts.

Unidentified speaker (UNKNOWN) ·22:49

How to verify: Check technical specifications for popular Snapdragon mobile processors (e.g., Snapdragon 8 Gen 1/2/3) on Qualcomm's website or tech review sites like AnandTech. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Raspberry Pis cost $100-200 and have 4 cores with 1-4 GB RAM.

Unidentified speaker (speaker_1) ·22:24

How to verify: Check current pricing and specs for Raspberry Pi 4/5 models on the official Raspberry Pi website or major retailers. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Luke's RTX 5090 was purchased for $2,000 a year ago and is now selling for $11,000 on eBay.

Unidentified speaker (UNKNOWN) ·27:38

How to verify: Check eBay sold listings for 'RTX 5090' to verify current market prices and search historical price trackers for GPU prices from 2026-2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Luke runs a trading card game simulator on his phone farm for three months.

Unidentified speaker (speaker_1) ·29:15

How to verify: Search for the specific trading card game simulator website or Luke's YouTube channel for documentation of this server project. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Samsung phones require manual power-on after battery removal, while Orbic Joy phones can boot automatically on AC power.

Unidentified speaker (speaker_1) ·11:13

How to verify: Check Android developer documentation or phone repair forums for settings related to 'Auto Power On' or 'Boot on Charge' for Samsung vs. Orbic devices. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 claims that GPU prices have risen over 500%, while phone prices have only risen about 5% (e.g., $1 on a $20 phone).

Unidentified speaker (speaker_1) ·30:05

How to verify: Check historical pricing data for consumer smartphones and high-end GPUs (e.g., NVIDIA RTX series) from retailers like Amazon or Newegg over the relevant period. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 states that phones are manufactured in the billions annually and recycled in similar volumes, whereas GPUs are not produced or recycled in comparable quantities.

Unidentified speaker (speaker_1) ·31:16

How to verify: Review industry reports from IDC, Gartner, or Statista on global smartphone shipment volumes and e-waste recycling statistics versus GPU production data from NVIDIA or AMD. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 asserts that a Red Magic phone with 16GB RAM can run a 30-billion parameter AI model at 30-40 tokens per second.

Unidentified speaker (speaker_1) ·33:47

How to verify: Search for technical benchmarks or blog posts from Red Magic or AI model developers (e.g., Llama, Mistral) detailing performance on mobile SoCs like Snapdragon. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 claims ChatGPT generates 1,000-2,000 tokens per second, compared to 30-40 on phones.

Unidentified speaker (UNKNOWN) ·34:07

How to verify: Check OpenAI's API documentation or independent benchmarks for GPT-4/GPT-3.5 token generation speeds on standard server infrastructure. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 states that web scraping companies use phone farms with SIMs to evade detection by services like Cloudflare.

Unidentified speaker (UNKNOWN) ·36:45

How to verify: Review Cloudflare's blog posts or security reports on bot management and anti-scraping technologies, and industry reports on proxy/phone farm usage. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 reports running 3,000 phones at his home with a 15-cent/kWh electricity rate, costing ~$400/month.

Unidentified speaker (speaker_1) ·43:28

How to verify: Calculate based on average power consumption of mobile phones (typically 5-10W) multiplied by 3,000 units and 730 hours/month, then compare with local utility rates. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 earned $45,000 worth of Night tokens from a gamified mining event for Midnight Network, which had a lockup period that reduced its value.

Unidentified speaker (speaker_1) ·45:37

How to verify: Check Midnight Network (Cardano ecosystem) announcements regarding the gamified mining event, token distribution, and lockup schedules; verify token price at the time of unlock. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Speaker 1 claims a game server running on one of his phones generates $500/month in recurring revenue.

Unidentified speaker (speaker_1) ·45:13

How to verify: This is a specific private business claim; verification would require access to the specific game server's financial records or public revenue disclosures if it is a known entity. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker 1 mentions that SIM farms are distinct from phone farms, with SIM farms using boards of SIMs for mass SMS, often leading to FBI raids.

Unidentified speaker (UNKNOWN) ·40:49

How to verify: Search news archives for FBI raids on SIM box operations or SMS fraud rings, particularly in New York, to confirm the distinction and legal risks. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Speaker 1 states that Ackerast is a significant revenue source for his operation, alongside mining Varius and Monero.

Unidentified speaker (speaker_1) ·44:16

How to verify: Check Ackerast's public documentation or user testimonials regarding revenue sharing or service models; verify mining pool statistics for Varius and Monero if publicly available. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Do your own research

  • Verify CellHasher's claimed shipment volumes (30,000 units) and network size (100,000 devices) against public records or independent audits.
  • Check current Varis token price and liquidity pool depth to assess realized value of mining rewards versus initial claims.
  • Review Google's USC research project details to confirm the specific use case (grading software) and compute requirements.
  • Analyze Midnight Network's token vesting schedules and historical price performance during lockup periods to evaluate gamified mining risks.
  • Compare electricity costs in the speaker's jurisdiction ($0.15/kWh) against local rates to determine profitability viability.
  • Investigate web scraping companies like Ackerast for current demand rates and contract terms for distributed phone compute.
  • Assess the technical barrier to entry for Linux-based phone mining versus managed solutions provided by CellHasher.
Full transcriptexpand

0:00Unidentified speaker (speaker_0): Are cell phones the best kept hardware secret out there that no one's talking about? In today's video, I sit down with Luke from Cellhasher to dive directly into this. All right, guys, I'm here with Luke from Cellhasher. How's it going?

0:16Unidentified speaker (UNKNOWN): It's

0:16Unidentified speaker (speaker_1): going pretty good. Thanks for having us today.

0:18Unidentified speaker (speaker_0): Absolutely, absolutely. So, as the title says, are cell phones the best kept hardware secret? So I think we need to kind of step back a little bit because you're like our our phone mining connoisseur here. You know, when did you actually get started tinkering in this space? And what ended up leading you to cell phones?

0:42Unidentified speaker (speaker_1): Yeah, uh, I've been in this space for a while, I would say, uh over a decade now, which is crazy. Um, the first initiative that I had with was just building computers, which learned into playing like games like World of Warcraft when I was like 13, 15, that range. And I wanted to buy in-game items. So to buy in-game items, I would go to these sites and you could pay with like crypto. At the time, it was like Litecoin and Bitcoin was like the main ones that you could buy with. So I was like, how do I get this? Like, what is Litecoin and Bitcoin? And led me into building computers, which uh took me on a path of development, um, developed for Litecoin as well as some things in Ethereum, like Ethereum Name Service, and a few other uh different dApps and stuff like that. But we're here now, we're mining with phones. Uh phones weren't always the uh go-to uh for a lot of us. Obviously, we went through FPGAs, see like the different CPUs GPUs to the GPUs to uh all kinds of different chipsets that people use today. And yeah, now we're on phones. Um I use phones for now since gosh, 20 about 2020, 2021. Um and it was all off a hunch of like

1:48Unidentified speaker (UNKNOWN): um

1:48Unidentified speaker (speaker_1): my power bill was slightly higher in the place that I was and I still wanted to mine at home. Um I could mine wherever and I didn't want to have to send uh just like little dinky hardware off to hosting companies. Um it just wasn't feasible. So all of the Bitcoin miners, sure, all of those really big ones that need you know a lot, a lot of power, send them off. And I was looking for different ways to, you know, mine to do different compute, because I'm not just a miner, I like to do other stuff and just overall a computer nerd in general. So yeah, we looked into phones because they were low wattage. I didn't know if I'd be able to power them without the battery, but obviously like all things I'm like, you're probably gonna be able to figure that out. So after some tinkering, we figured out how to power them without the batteries and then without the screens. So that became more compact and just the boards. And yeah, that's kind of where we're at today. It started with a hundred. I bought a hundred kind of on eBay randomly. I'm like, you know, I'm gonna, if I don't buy a hundred, I'm not gonna make this work. So I bought a hundred and um, yeah, that took into like a long three. month process of figuring out how the heck I'm gonna get these to run because I wasn't necessarily too hardware focused. Um I went to school for hardware um rather than the networking side of IT. But they never teach you about like the really small electronics other than just like your Arduino boards or Arduino boards as some people would

3:08pronounce it. And yeah, so we're here now with the phones.

3:12Unidentified speaker (speaker_0): Love it. Love it. So I think we need to step back in time. a tiny bit there. So sure. So you're a glorified gold farmer, is what you're saying from World of Warcraft back in the day. Is that what you're telling me?

3:21Unidentified speaker (UNKNOWN): Oh

3:21Unidentified speaker (speaker_1): my gosh. Yeah. And the uh unfortunate bit is World of Warcraft Forever just came out a couple of days ago. And

3:28Unidentified speaker (UNKNOWN): uh

3:28Unidentified speaker (speaker_1): a lot of my a lot of my old buds are online, and I'm like, oh man, I will play, but I just I don't know if I can commit the time to it anymore. So yeah,

3:37Unidentified speaker (speaker_0): definitely hooks you in. I I got into World of Warcraft much later than most people did. Uh my first experience was the night like Burning Crusade came out. Um I stepped in right about right around the South Park episode. The South Park episode introduced me. Um and then Burning Crusade came out a few weeks later. But uh that's absolutely awesome to hear uh that like that's where it started and that where your that's where your introduction was to crypto because then you know it was gold buying gold wasn't very it wasn't legitimized right it was kind of through these shady websites yeah hence where some of these cryptocurrencies came in so yeah you were you were quite back in the day for sure

4:15Unidentified speaker (UNKNOWN): so

4:15Unidentified speaker (speaker_0): taking a look at that you know um you started right around like the twenty twenty time period that you were alluding to with some of these uh other projects that you were getting into did you jump into GPU mining at all? Or did you kind of scoot around that and go go more into this cell phone interest?

4:35Unidentified speaker (speaker_1): Yeah, so I did uh I did do the GPU mining, but like I at like as like a 15, 16 year old, I didn't have um a lot of cash on hand. I did work other jobs so that I could buy it, but I had like the what is it, the RX uh gosh, fifty 5080s, eighties, forty 4080s. eighties. Um

4:52Unidentified speaker (UNKNOWN): probably the 580,

4:53Unidentified speaker (speaker_0): probably like the 4070s. Yeah,

4:54Unidentified speaker (UNKNOWN): 580s, yeah, yeah.

4:55Unidentified speaker (speaker_1): Like just really like cards from I think that's like 2017, 2018 era. Um, I had a few of those. Um, I did that, and then I did, I was able to expand into the like RTX series as they were coming out, the 1050s and stuff like that. But I never had anything crazy like a lot of people might think. Um, but it was it was more than a hobby. Like I had, I think, 20 GPUs at one point um when I was like 17, 18, but it wasn't anything crazy. So I can't say it was big, but it was it was a nice operation at the time because it's still through college a little bit. Like I had that off in the room in college was free electricity, you know, to an appointment. So I used Use that as much as I possibly could. And, you know, that worked pretty well.

5:37Unidentified speaker (speaker_0): Man, I wish I mean I'm older than you, but I wish I knew like what I know now and like in college and taking advantage of that, because that man, you could you could have made some some good money then. But okay, so we have these cell phones, right? And I think people naturally write them off, uh, is what I find, where it's like, oh, that's stupid. Uh, you're not making much type of deal. And I think you hit the nail on the head earlier where you were like, the one. watt usage is barely anything. Um, and you can really do a lot with it. But something you did mention was like you were gonna come out of the gate buying a hundred. And I think that really leans into, and you can kind of explain it because I think you not only deal with this on scale, but also from like a manufacturing side of like, yeah, one or two, it's not gonna going to be much, but when you're talking a hundred and you start to scale, you're actually talking about some real revenue there. Today's podcast is sponsored by ASIC Marketplace.com. I've actually been working with Rain and the team at ASIC Marketplace for several years now, and I can personally highly recommend them. In fact, I purchased my first Bitmain Antminer KS5 Pro Caspa Casper miner from them. Are you looking for some serious hash rate? Go over and check out the BitDeer A2 Pro Bitcoin miner or the Bitmain Amp miner L11 for

6:56Dogecoin mining. I'm also eyeing up not yet released the BitDeer DL1 Air. Guys, this thing pushes 25 gigahash for Dogecoin mining and the BitDeer A4 Pro Bitcoin miner packing a whopping 336 terahash. ASIC Marketplace will be my go-to when these miners come out. And for home miners, they've got you covered with loads of the latest generation solo Bitcoin miners. Listen, go over, visit ASICMarketplace.com and use the link directly down below and use the checkout code Everyday Day Miner. minor. That's right. Everyday Every day miner. minor. One word to save big on every every ASIC purchase.

7:41Unidentified speaker (speaker_1): Yeah. The the thought process on the hundred originally was

7:45Unidentified speaker (UNKNOWN): um,

7:46Unidentified speaker (speaker_1): yeah, two's not gonna cut it. Like I'm not gonna make anything with two. Like when I ran just the like initial numbers, like all of us do, we all crunch numbers and figure out, you know, how's this gonna work for us? You know, two, two wasn't gonna cut it. Two would like, you know, by definition, get us to the point where we need to eventually. but just not at the not at the time that we have available. So a hundred was more my commitment of like, okay, um, if I'm gonna do this and spend time on it, I wanna make sure that I get something out of it. So a hundred was my number where I felt like the time that I would put into it, I would get out of it. So, um, and sure enough, um, that worked. At the time,

8:24Unidentified speaker (UNKNOWN): uh,

8:24Unidentified speaker (speaker_1): Varus was like the big one um for. for you know phone mining per se. I was gonna use them for other projects. Um, like you could do it like running you just like your own PyNets and other hardware servers. You can run ad blockers on your network, which uh Pi Hole is one of those services that you can put on a phone and it uh blocks ads through your network. So those were kind of like the use cases. I was like, all right, I'm gonna mine just a little bit. I'm gonna figure out how to do this because. I know nobody's really doing this. Um and I think it would just be fun, kind of like uh in a way, like a little bit of a flex. Because on on X, you uh as uh my generation, you you want to show stuff that you're doing and you want to do stuff differently. Um So that was my nerd flex of all right, I'm gonna set up all of these different phones and uh get them working. And it was a struggle at first uh because there's no like documentation. nobody hands off and you can't just say say hey google how can I take apart your phone and uh do it in this fashion that they don't do that. I mean Google doesn't even do that and we can get into that uh later on but um

9:28Unidentified speaker (UNKNOWN): yeah

9:28Unidentified speaker (speaker_1): the the hunter phones on Varys right away was I had just an old desktop Top um chassis and I had just bought stuff on Amazon that had like slots and I just did the really easy route before I went uh to try to make anything uh substantial. And it was all kinds of wires and plugs, uh plugging everywhere. Um, kind of a mess, to say the least. Um, the first 20 that I got up worked pretty good, and I saw um some good returns with various. I I can't remember exactly what all 20 were making at the time, but it was like two, three dollars. a day, which I was happy with on phones that I had purchased for like 20 bucks um per phone, roughly, uh was the case. And they those were uh Orbic Joys and AO3S. And I have since moved on from those. Um Orbic Joys, I still have some of those and I still use some of those. The AO3S O3S were just a little harder to uh manage software software-wise. wise. Um they didn't have some features where as I found out, as you kind of put these phones together, you take the batteries out of them, they still need to think that they have a battery in order to function. That's just how the software is written. So if you don't like root the device, which most of the times you can't find publicly available resources to root these things, Um, you're left stuck with the hardware and

10:48firmware that's already on the device from the manufacturer.

10:51Unidentified speaker (UNKNOWN): Um

10:51Unidentified speaker (speaker_1): that being said, Samsung

10:53Unidentified speaker (UNKNOWN): uh

10:53Unidentified speaker (speaker_1): was pretty locked down and I didn't have time to really mess with it. So when you take the batteries out, it was really hard to get them the power on um or power on without you manually pressing a button. And when you get to a hundred phones, you don't really want to manually press a button all the time. Like it's doable, and I would have done it, and I did do it for a period of time. Um, but I just it like it wasn't at scale. And then Orbic Joyce came along where you could turn on this one setting within Android where when it is plugged into power, um AC power, it will boot. Um before they would boot to a charge screen and then it would shut off. And that's kind of like the Samsung thing and some of the other ones. And since there's other models of The Samsung that do allow this stuff, um, which we do kind of offer um but nonetheless it hopped into that. And then I went full force into Orbic Joys with a hundred of those, uh, set those up. And I I mined Varus for quite a long time, about

11:48Unidentified speaker (UNKNOWN): a

11:48Unidentified speaker (speaker_1): year and a half, two years, kind of stacking off of a hundred phones. Um, and then I had scaled more, like I had put more phones online as I had seen the returns kind of coming in. And thankfully, Varus at the time, I would say it's it's very lucky. The cycle was doing. good in crypto. And Varus went up like 600, 1200% since I had gone in. And I haven't really sold anything. And they had a good community where they had a bunch of like liquidity within bridges, not necessarily exchanges. So it was almost like this secret cheat code where you could go to the liquidity pool and swap it for Bitcoin. And I was able to walk away with quite a bit of money off just essentially a 100 hundred phones plus plus some. But has since scaled. Um, I'm at my house, I'm a little over like 3,000 devices, um, phones. Um, we have other places that we run stuff now, but um yeah, it's pretty crazy. I think this next year, um, or even during the winter, I might hook up some solar now to it. Uh

12:46Unidentified speaker (UNKNOWN): not

12:46Unidentified speaker (speaker_1): that I need to, but just for the the proof of point, where these can run off a 200 watts. watt solar panel like one box and then you're

12:53Unidentified speaker (speaker_0): yeah

12:53Unidentified speaker (speaker_1): uh you're mining free at that point. So

12:56Unidentified speaker (speaker_0): yeah, yeah, that's incredible. That's definitely incredible. So uh man, 3,000 phones, man. I'm like a hundred almost a hundred. I'm like, yeah. So okay,

13:04Unidentified speaker (UNKNOWN): so

13:05Unidentified speaker (speaker_0): I think that's really important. So you definitely went down this rabbit hole, and we're gonna talk more about like actual revenue there because I think that's important. But when did this go from okay, this is a high hobby or this is a small business at home of you know um mining yourself to like, hey, I want to start a company, and that's really where Cellhasher came into things.

13:26Unidentified speaker (speaker_1): Yeah, I had a lot of I posted a few videos on YouTube um kind of like documenting the process, and it breaks down all of the numbers I was making back then. So you guys can all go back and watch that if you want, and you can. kind of uh see at the point uh what price Varus was when I was mining. And it was very low compared to where it was uh then in the future. Um

13:49Unidentified speaker (UNKNOWN): but

13:49Unidentified speaker (speaker_1): yeah, I

13:50Unidentified speaker (UNKNOWN): I

13:50Unidentified speaker (speaker_1): had gone in um with all of these devices. Um basically

13:56Unidentified speaker (UNKNOWN): I

13:57Unidentified speaker (speaker_1): went through the first 100 to scale, and then I realized at that point, the money that was coming in, I could buy a phone roughly like every couple days. Um, so then I just I took more money. I didn't take the money out, I took more money, and then I started buying more funds, more phones for that. And people started commenting on these videos and kind of reaching out to me. Hey, I kind of want to do this too. Um, so I started helping them out on the side, and then we got into like, okay, we should probably make some. like other designs for these because it's just not viable. I was showing everybody like the computer setup I had and this like big phone rack. And I think retro Mike was getting into it and he was doing connects at the time, which you know everybody has some people were 3D printing stuff at the time. So there was there was a group of us that were kind of already like um talk talking more about how we can like do this. at scale, which at that point that's how like Cellhasher came around. Is um I reached out to a few manufacturers that you know did metal fabrication. Um, and

15:00there's really none around here, sadly, in the United States, that could that would do stuff small enough for like an affordable price. So I had to go overseas with it, which wasn't the worst case, but you know, waiting a couple months for like the first prototypes was I was so eager to get it. in my hands. But um I do like I do have a background in in some engineering. So we were able to do some of the first uh PCB boards that had it really basic stuff, no switches, anything. It was basically just USB 2.0 because it's cheaper than 3.0. So USB 2.0, power delivery, um, and then just connection to the computer. So another another data connection. And it was it was just what it needed to be. There was no switches. But then we learned really quick off the first one that all of these different phones require different voltages. No phone model or different models, excuse me, operate really at the same one. Some will run generally around five volts, but some might run like 4.5 five volts or 5.1 volts, depending on you know the year they were made, uh the type of architecture they carry.

16:03Um but yeah regardless back to like the the business side of things I genuinely wanted to like help people. And obviously like at the time I was thinking you know like if other people want this like this can also fund my phone farm as well to scale and do that. Um which it should most definitely has. I would say though

16:22Unidentified speaker (UNKNOWN): uh

16:23Unidentified speaker (speaker_1): this is this is going gonna to sound probably shocking, but I feel like if I would have focused on just my phone farm, I probably could have scaled that and made a lot more money than actually starting a business, uh, believe it or not. So I think about

16:39Unidentified speaker (speaker_0): people and you know, spread the word and you know, oh, how dare you, you know? But no, I I totally get that. So, so jumping ahead here a little bit uh with Cellhasher, you know, where is Cellhasher today? You know, what does this look like? Um, give us a rough number. Like how many cell phones you estimate have shipped out? Oh

16:57Unidentified speaker (speaker_1): gosh.

16:57Unidentified speaker (speaker_0): Um what

16:58Unidentified speaker (UNKNOWN): do you

16:58Unidentified speaker (speaker_0): guys what are you guys focusing on now versus where you were? Because I feel like Cellhasher has really evolved over the years, even since I've met you.

17:06Unidentified speaker (speaker_1): Oh my gosh. Yeah. Yeah. I mean, we weren't uh we weren't anything special to start. We were literally just, you know, here's how you can do this for. yourself. And here's your DIY options, which then led into us actually doing the work for everybody, which was a blessing and a curse at the same time. The the load of work is insane to this day. I probably have taken apart way too many phones um than I should have in anybody's lifetime. Um But yeah, I mean

17:37Unidentified speaker (UNKNOWN): as

17:37Unidentified speaker (speaker_1): it's evolved today, I think we probably through the last I mean we've been around for it's 2026 four years now. Um we've probably shipped off maybe just under 30,000 phones. Um so or I mean it would be more if we account like the people doing it themselves, like DIY units where you don't get stuff with phones. So it's significant. Um, I would say at that point, probably our devices now hold around 100,000 phones, I would say. Um, and that's just including like consumer. And then we have some like enterprise uh customers now.

18:17Unidentified speaker (UNKNOWN): Um

18:18Unidentified speaker (speaker_1): we just and I can I can say this, but we we did a deal with Google recently, which is huge. I I count that as like a a big success. Um hopefully they come back for more. Uh, but we'll we'll we'll see. And that's still like ongoing. Um, Google did a research project, and I can kind of talk on this, um, or excuse me, they funded a research project from USC, um university in California there, on uh basically using, you know, this recycled compute to run college grading software and run other computing programs. So that was the specific use case there was running grading software for USC and other small tasks where right now in the AI craze, some of the really heavy jobs they want to keep obviously for like the AI and they want to run their computers and servers to facilitate that. And that's good. That's okay. But that leaves a really big gap for a lot of small compute. And the the thing with like running small compute on a GPU is good, but it's the time of like switching to that, turning off what it was already working on and doing that and then hopping back on, sometimes it doesn't make sense. So they were bridging that gap of okay, like there's is I th I don't know the the number. It was in the billions of phones that are recycled every year.

19:36Unidentified speaker (UNKNOWN): Um

19:36Unidentified speaker (speaker_1): and that's and that's just e e-waste waste. Um to an extent, where they're they're taken off, they're shipped, and yeah, they do get some get recycled, and people like me and other people do find ways to use them. Um, but it's at such a level of um e-waste that is like uh just sitting there um being useless. So they were trying to find a way and they were they uh did some funding and granted USC uh and then the person graduated is now with google um and they're doing their own farm now so they're they're um putting together a farm of around 2000 phones to start um for various things they're doing the pixel route imagine that google's doing pixels which is something something we haven't done um we didn't we haven't messed with pixels most because you know google locks down their stuff so you know we but we but when you're working with Google, uh they can do the stuff they need to. They can take the software off. They can make it you know work perfectly for what they're doing. So yeah, we made some custom chassis. They're like 60

20:36phones in each chassis so they can fit into their servers. It's it's pretty cool. We have that will be going on now for like the next you know couple a months. We still have that whole integration going with them. So that was that was our most recent. So to say that it's been viable even for Google to at least, you know, test the waters and give it a try, um, I think that itself serves a point. And like, you know, this compute is definitely useful still. So

21:02Unidentified speaker (speaker_0): yeah, I think it's interesting because like the world of my audience and where I come from, you know, we look at like GPU mining, right? And like we did that back in the Ethereum time period. And now it's becoming more and more popular with Perl Pearl and Quantus and some of these other ones there. But a lot of these are starting in some way, shape, or form to all come back to AI. Like it's all coming back to AI in one in one way or another. And it's interesting because like where my audience might think, okay, like one phone, as we talked about the top of this, isn't really going to make a difference. And I think everyone can agree about that. But when you look at these at scale, you could get some decent amount of compute with. them, which, as Google is looking at and doing, which is a great example, especially when the cost for hardware is through the roof. Demand is massive, supply is very minimal when it comes down to GPUs. And as you had said, there's so many phones being recycled every year. It's just it's like untapped compute, You're just ready to sit there. And I think, correct me if I'm wrong, it all comes down to that snapdragon process. processor. Is that correct?

22:04Unidentified speaker (speaker_1): Yeah. And there there's tons of different processors and phones uh today. But like, yeah, um, the processor, I I think about it like this. Here's maybe a better explanation for people is um a lot of people have have tried like the Raspberry Pi things and they set up Raspberry Pis that do maybe like some CPU mining and stuff like that. Well, your Raspberry Pi will cost you like $100 to $200 these days. Um And you get like four cores and you get like maybe a gigabyte or two gigabytes, maybe four gigabytes of RAM in most of those, or like rock chips. A lot of people are familiar with those because people did those with FPGAs and all kinds of things, all kinds of things like that. Whereas your phones, yeah, you have the Snapdragon chips, which are made to be extremely, you know, performance-heavy, performance heavy uh, but on efficiency scale. So you have

22:49Unidentified speaker (UNKNOWN): the

22:50Unidentified speaker (speaker_1): majority of phones today have eight cores. Um they run around like you know two gigahertz, some get up to three and four gigahertz now. Um, which doesn't seem like anything special, but uh the compute is like you're running at less than five watts. Um so you're getting eight cores. Um most phones will have three to four gigabytes of RAM, like the really cheap ones, like twenty $20 dollar prepaid phones. Um and then you get into like you know, consumer phones, which now have like eight and twelve gigabytes of RAM, along with those cores. And some of those chipsets will have, even some of the Snapdragons will have NPUs, which is something that, you know, uh some of the like new computers that are coming out that are like really small and form factor that can run AI like on your desk, all of them are running like NPUs, MPUs, neural processing units. So the fact that the chips are still expanding for phones to give us more performance, and the MPUs will work more efficiently with AI. Now, a lot of them are firmware based, so the MPUs today are not as usable as people may think, but they they are to an extent. Um They're kind of firmware locked to like you know, Google's own system of how you work on. a model or Samsung's own system of how they work models on their phone. Um, you can technically unlock

24:09them. Um whether that's like I I don't know the whole if uh you're allowed to do that. I'm assuming you're not allowed to do that. Um, but you can unlock them, just like you know, the CMP um unlock that happened, where they could access more memory on. that chip. Um that was crazy. And this is roughly the same thing where if you can access the MPU, you can now push like AI workloads through the MPU. You can still use the onboard RAM of the phone and you can get better performance. Now you could always run an AI model with a CPU and the RAM that's that on the on the system already. Um you're just not gonna run anything like special, like anything crazy. Um, like if you're thinking you're gonna run some of the latest like consumer models, um that that's a little bit of a waste of time. Now, where it is going to pay off for people is, and what we're seeing a lot more of is today, a lot of the models we have, um, they're a generalized model, but then they're they are using other models as like um subset agents. So they're they're spawning an agent that is specialized in this task. Well this task has been trained as its own model. So it has like a lower weight parameters is the term that they use. And that can fit on a phone. So if you have, you know, for example, and what something I would love to do in the

25:29future and maybe make some videos on is with a thousand phones, let's set up a thousand different agents that are particularly good at you know one thing where one is just the coding agent, but it's only the coding agent in Python. One is just the coding agent where this is only the coded coding agent in Rust. Um, one is your writing agent, one is your um scripting, engineering, uh workflow uh manager, right? So you can specialize and do all that with small models across the phones, and then they'll run decently efficient um, in in AI for that matter. Um, but nonetheless. yeah, your the chipsets, the Snapdragons are very powerful. Um, they're way more powerful than a Raspberry Pi in terms of compute. So the people who Raspberry Pi mined Varis for a long time, they did really well. But then when the phones came on, um, they still ran it and it was easy to manage. Um, but the phones blew every everybody out of the water in terms of efficiency with the compute. And there's plenty of stuff you can do and mine uh with phones. So

26:27Unidentified speaker (UNKNOWN): yeah,

26:28Unidentified speaker (speaker_0): well I want to get into that in a second. I think, you know, a first question I have though is like right now we're seeing where compute is coming to a head. And and and and I don't see AI, I don't see that going anywhere. You know, we only see compute and AI demand going up. But what I mean by that is like our cost for GPUs, our cost for memory is is at an all-time high. And that being said, we don't see supply going getting any better, right? It's not like all sudden NVIDIA is like we launched 10 different foundries. Like that's not happening. Like these memory manufacturers aren't doing the same thing. And for the longest time, cell phones have kind of been this great best kept secret, especially for home miners and tinkers, because as you said, it a relatively massive horsepower for very, very low watts. And with Google now stepping into that space and looking at this, of like, all right, we're seeing we're gonna need more compute, we might need to hand it off on that side. Um, do you think there's gonna be a time here coming soon where all of a

27:28sudden cell phones are gonna blow up as another avenue? Because hey, I can't get my hands on any cards without spending like a great example my 5090, I have one of them per for

27:38Unidentified speaker (UNKNOWN): se.

27:38Unidentified speaker (speaker_0): sale on eBay for eleven thousand dollars right now. Imagine how many cell phones you could buy. Yeah, yeah. Imagine how many cell phones you could buy for eleven thousand dollars. And I bought it for two thousand a year ago. So I guess my thought is I imagine that you're sitting in this situation right now and your wheels are constantly turning as to like what's coming next.

27:59Unidentified speaker (UNKNOWN): Yeah,

28:00Unidentified speaker (speaker_1): that's fair. I I have some fifty nineties at home, so now I'm gonna go give it a look when I get back

28:04Unidentified speaker (UNKNOWN): and

28:04Unidentified speaker (speaker_1): see if I really need them or not. Um but yeah, like you said, like I think of that. And if if I could sell one of my fifty nineties for eleven thousand dollars, I could probably get a setup of like, oh,

28:16Unidentified speaker (UNKNOWN): you

28:16Unidentified speaker (speaker_1): know, two hundred to four hundred phones, depending on the range, which is a big increase um for anybody. And at that point, is it worth it compared to the fifty 5090? ninety? Um in terms of what it could produce in revenue? I would say it's I would say it's decently close. I uh with with quotations, like uh the GPUs will probably always have more like jobs available to them currently. Uh whereas the phones, you're gonna you're gonna just have to kind of stick around with what's always been around um or you know go and find other ways to do things. Um that like that being said, uh like currently I have uh another like side business that I'm running on the phones because I'm gonna make a whole process on how people can like do this stuff. And to get into that a little bit is like I'm running a game server um that I made a website of of kind of a trading card game. And it's a simulator. And people have been playing on the simulator and the simulator's been earning money. Well, what they don't know. is the simulator's been running on phones for the last like three months now. So uh it's it's something to kind of show that, you know, this isn't just like mining focused. Like you're the home my it's for like the, you know, the the home lab people as well. You can run your servers, you can run your stuff on it, and you can be efficient. Like I'm not paying heavily

29:35on costs cost for those servers, but yeah, like you said, I mean the the the back to kind of the whole premise, the the compute that you get for it, uh it's just it's mind-boggling. The the comparative to the the GPUs, I still think like obviously the GPUs will always have a place um going forward with memory not coming down um in price, with storage not coming down in price, all all compute um from all different angles. The phone set. phones have relatively sat. Um they've gone up slightly, but nothing nothing like that. Like maybe like one dollar on a twenty dollar phone. So, you know, five percent um compared to like, you know, your GPU at that point, like you're well over five hundred percent up on that. So um

30:20Unidentified speaker (UNKNOWN): to

30:21Unidentified speaker (speaker_1): to say the least, I hope not. I think with

30:24Unidentified speaker (UNKNOWN): I

30:25Unidentified speaker (speaker_1): don't think it will happen for phones. Um and I say that I say that pretty confidently. And the reason I say that is because the phones are manufactured differently than GPUs are manufactured. Um if they're a lot easier to produce,

30:41Unidentified speaker (UNKNOWN): um,

30:42Unidentified speaker (speaker_1): they're a lot cheaper to make. Um yes, parts like memory and storage are getting more expensive. Um, but I think there's enough phones and supply and in just,

30:55Unidentified speaker (UNKNOWN): you

30:56Unidentified speaker (speaker_1): know, kind of circulation that we will have it throughout this time where everything's really expensive before new things come along to fill those gaps. Because eventually, like, I I can't see a world where this is how this lives, where everything is, you know, s rocket sky high price because memory's up, you know, uh storage is up, all kinds of things are up. Um, somebody's gonna come in and fill that void and make and make a solution or make an upgrade to like, hey, we don't need to use memory in this fashion. Here's a new upgraded way to use memory. Hey, we don't need to use storage in this fashion. We now have instead of H H Ds or HDDs, SSDs, we now have, I don't know, um, LDDs, whatever that whatever that is. Um so I think phones are at a at a level of like, I mean, it's in the billions that are recycled every year. I don't know if billions of GPUs are being produced every year or recycled every year. Whereas you have Samsung, you have Apple, you have Motorola, you have Google all making

31:56cell phones. And they're all competing for you know billions of users because in in the world of people, everybody has a cell phone on their hand rather than a GPU. Um it's just more of an everyday device. So I don't think that they're gonna,

32:12Unidentified speaker (UNKNOWN): I

32:12Unidentified speaker (speaker_1): don't think we'll have a problem getting them. Now, yes, we have seen initial like buyouts of stuff, but then more just come in and get recycled. I think maybe where we'll see maybe some like slowness is the recycling portion. Um, where people do hold their phones back a few more years um and they use it while they have it because like, yeah, they want to save a few dollars. They don't want to buy the newest phone, or they realize they can use it for compute um um in in other forms and other fashions.

32:42Unidentified speaker (UNKNOWN): Um,

32:43Unidentified speaker (speaker_1): but yeah, I don't I don't think uh the phones will have it as bad as GPUs, which is another reason why like you know, might as well lean into it now. while that is around. So yeah.

32:53Unidentified speaker (speaker_0): Yeah. Yeah. No, it's interesting. So for my audience, you know, I've heard just from chatting with you over the years, all these different ways that you can use phones. You know, we've talked specifically about mining and you talked about Varys there. Um, some people are using um phones for like Monero. Um, what you know, what other avenues other than okay, we talked about phone mining. So like using them for crypto mining, lots of different platforms out there. On the AI side, how are you seeing them used right now? What are some great examples that you see out there from most of the customers of Cellhasher?

33:27Unidentified speaker (speaker_1): Yeah, a lot of the AI side, um, you can generalize it into two categories. One is like a slight AI side and one is a full AI side. The full AI side is running the really small models. on the phones. Um so like for example, we were able to run a 30 billion parameter model um on a red magic phone, which is like a top-of-the-line phone. Um it's got an MPU, it's got a GPU, or excuse me, not a GPU, it's got an MPU, it's got the the latest Snapdragon, um, but then it's got also uh 16 gigabytes of RAM in it. Um and there's some models that have 24 gigabytes of RAM phones now, too, which is absolutely a absurd, but um, I'll take them. They they have

34:07Unidentified speaker (UNKNOWN): uh

34:07Unidentified speaker (speaker_1): we're getting around like 30 to 40 uh tokens per second throughput um generation. So it's it's coming back like relatively usable. Like 30 tokens a second is very slow compared to like what you get back from ChatGPT. ChatGPT is probably running around a thousand to two thousand tokens a second. There's probably people who know those numbers um on like the newer models. Um but that's kind of the interface you're used you're used to um using. Well if you can put them on the phones and you can get the you know 30 tokens per second and you have a hundred of them, it doesn't necessarily matter. You can walk away for five minutes and come back and have all of your stuff set up and resolved. So yeah, some people are doing it the route of um

34:47Unidentified speaker (UNKNOWN): do

34:47Unidentified speaker (speaker_1): using the small agents on the phone and then prompting a like a management agent that goes and distributes to all of the phones the different agents uh that need to run the compute. So yeah, people are using it for everyday like AI random randomness, whether that's like uh image generation, to an extent, not so much. Um, but like the coding portion we we've seen people use, and I've used phones to code. I I don't use them. as much as people would think because there's not necessarily I don't have a good reason to use it for the code quite yet. Um but some people might. Like it's not just uh like here in the US we have access to a lot more hardware, I would say, than people in like maybe um not to like like in East Asia or any of those other places where um they don't have access to compute. So like they're gonna sit there for five minutes and wait for their coding session or their coding agent to finish if they can get it for free. Right. So

35:47Unidentified speaker (UNKNOWN): and

35:48Unidentified speaker (speaker_1): that's and that's being done. Like uh like having the you know the North American eyes kind of uh blind to that is like it's a big thing over there in East Asia is like people are using their phones to do all of these different things with AI agents. Um, and they're doing other things. Like, for example, what was the one? It was like Axi Infinity. There was a crypto project like on Ethereum a long time ago. And the Philippines had people that became millionaires in US dollars because of that game, but they never played it really on the computer. It was always like on their phones, and they would have multiple phones. out doing this all the time. So that's the access to hardware that they have. And that's kind of what we're seeing on the AI thing for that. What we're seeing inside of like the US and Europe is a lot of uh data and web scraping. So they're using devices like phones because they simulate real users. And that's

36:45Unidentified speaker (UNKNOWN): like

36:45Unidentified speaker (speaker_1): to be an extent, it could be a terms of service thing for a lot of things. So like people watch out for those things. but it's it's really hard to like stop. Like you're not gonna it's you're not gonna stop a a phone farm that's set up well enough. Um if you have sims which you can do in all of these phones, essentially, that's that's one user. So a lot of these like web scraping companies are using phones to do this. Um Acarast uses uses all these phones, and I would say their some of their big customers are probably web web scraping companies. Um like for certain I'm almost positive that they're web scraping companies based on kind of the jobs that I can see getting pushed through my network because I also run Ackerast. But like

37:32Unidentified speaker (UNKNOWN): the

37:32Unidentified speaker (speaker_1): the AI craze is using all of that data, obviously, um all of the web scraping data because they need to train their models. So anything that it can learn or pick up on is going to be used. So the phones don't get picked up as like normal computer bots as much. They can, but uh things like Cloudflare, which is exceptional at their bot detection and having rules for like um web scraping and all of those other things, um, I'm sure they have a good way to stop traditional phones. and to do that, but like from the web scraping industry, um, and if you're doing it legally, uh, it's the phones are the way to go. And that's how people have done it for honestly like a decade now. Um, well beyond me, is I found out that in Asia, a lot of these web scraping companies and even social media companies uh and management companies for social media platforms would all be using phone. farms. Of

38:28Unidentified speaker (UNKNOWN): course, theirs

38:29Unidentified speaker (speaker_1): also looked like how retro mics and mine started where it's all plugged in to different phones and they have this one guy clicking on everything. And uh that's not feasible, but like to an extent they've been doing it for a long time. But yeah, that's the other.

38:41Unidentified speaker (UNKNOWN): So

38:41Unidentified speaker (speaker_1): that's the AI light side of it, and the AI heavy side of it is the models. The AI light side is the scraping data, so that you can train the models. But yeah.

38:50Unidentified speaker (UNKNOWN): Yeah.

38:50Unidentified speaker (speaker_0): So we talked about crypto mining. We talked about AI and compute. Um, I've heard you know crazy stories from you. Uh, and I don't know if these are still the case, but um, there's a lot of either advertising or click farming or other avenues that cell phones are being used for. Is that as popular still today um that you find from just your side of the world as a uh distributor with the cell phones and all?

39:13Unidentified speaker (speaker_1): Yeah, the the use of cases that we get, I see a lot of,

39:19Unidentified speaker (UNKNOWN): we

39:19Unidentified speaker (speaker_1): see a lot of uh it's gonna sound crazy, but like uh click farming is a big one, right? So there's like four different genres of click farming. One is gonna be media click farming, the other is gonna be um ad survey click farming, uh, and then the other one will be ad watching click farming. So there's a bunch of different programs where if you watch ads, you can get paid, right? If you take surveys, you can get paid. Now that's grueling work for anybody. But if you set up a phone farm and now you put a small agent on the phone that can see the phone screen and interact with it, you have now made a passive ad watching uh click revenue um phone farm Um that does it all passively, um, which is again extremely hard to stop. So yeah, we do see clients that do that. Um one big client um of ours does uh proxying, so it's uh they they service data. So people will come through and that's like that's another thing that's like slightly different that we haven't mentioned, which is um you're paying for like a VPN, right? Uh the VPN goes through um a tunnel, it hits a proxy, and it kind of obscures your IP address. Well, phones do phones can do that. So you set up the phones as proxies, and that's kind of what some people do. So they'll

40:33Unidentified speaker (UNKNOWN): use

40:33Unidentified speaker (speaker_1): SIMs, of course. SIMS is like the big one. And I I want to be careful with what I say here because it's not a sim farm. SIM farms traditionally are set up to take down or to mass send text messages, which is uh

40:49Unidentified speaker (UNKNOWN): to

40:50Unidentified speaker (speaker_1): my standpoint, it's not um something you're supposed to do. Um and people do do this, and uh people have uh gotten stuff and had like the FBI raid them for this. Like there was a big operation in New York where um it was all SIMs, but most of the SIM farms don't include phones. Um they're big boards and they're like just SIMs that you stick into a board. Um, and that's the SIM farm. So it's it's completely different. This looks level of proxying is the legal proxying, it's the uh VPM proxying. So we got people doing that. Um, we have people uh doing uh in the music industry, there's a lot of uh producers, and I'm sure it's in every industry that will use phones to basically view viewbot their um clients' socials, right? Get their songs playing into more people, into more playlists, and that's what they're they do. Um that's that's to them. They can do what they want to do with that stuff. Um, I can't say we like facilitate that because we don't directly advertise that you're doing that, but sometimes you get people that come through our like uh support channels and they're asking those

41:54questions. Um, like, can this do this? And we're like, yeah, we can't confirm that it can do this. Um, so you have to like, you know, find this out on your own. But yeah the from the mining to the click farms um to running kind of like what I explained is like actually running a business or having your website um where you're say you throw up a a hobbyist miner website where you people can buy all of your you know hardware over over the time you could run that from a phone at five watts and and that could handle everything. Um So there's ways to generate revenue um on that side. Um yeah, I mean anything that you can do with a computer,

42:31Unidentified speaker (UNKNOWN): and

42:31Unidentified speaker (speaker_1): there's plenty of ways to make money with a computer. You can do it on a phone, um, just at a more efficient scale. So let your mind be creative um in that. And there's endless different different ways that you can make money.

42:43Unidentified speaker (speaker_0): Yeah, I mean, it's absolutely crazy to hear, and you can absolutely get creative. I mean, my mind's just turning on like all the different ways that you can utilize this. And it's it seems like the over the overall theme is like you know, it's not one, it's like massive amounts and collectives. So, like, you're not running a lot of this off of one phone, like your game servers or your web servers or whatever. You know, you really got to get a number of them together to do this. Um, and that's kind of where my mindset is with uh my cell phones that I've started diversifying with is like one or two or three isn't gonna do it, but once you start to have several hundred of them, that's when you can actually see some real revenue because you're you're outside of your upfront cost, your electric usage is let's be honest, it's close to nothing compared to what we're used to in the industry.

43:28Unidentified speaker (speaker_1): Yeah, it's it's crazy. Running 3,000 phones um at my house where I have just over a 15 fifteen cent a kilowatt hour rate. Um, I think our last bill, and obviously it's grouped in with a bunch of other stuff,

43:40Unidentified speaker (UNKNOWN): but yeah, um,

43:41Unidentified speaker (speaker_1): I've kind of sorted the phones out to be around four hundred dollars, but that's three thousand phones. Whereas if

43:47Unidentified speaker (UNKNOWN): you're

43:47Unidentified speaker (speaker_1): I was doing anything of scale um with other things, like I'd be in the thousands of dollars. And I am with like ASICs already at other places.

43:57Unidentified speaker (UNKNOWN): Yep. Nope.

43:57Unidentified speaker (speaker_1): That's a thousand, a couple thousand dollars a month already, but uh that's just the the life we live as as miners. uh to say the least. But yeah, the phones, uh the three thousand phones probably pull in average on a good month. Um we're doing absolutely nothing, right around six thousand, seven thousand.

44:15Unidentified speaker (UNKNOWN): Good

44:15Unidentified speaker (speaker_0): for you.

44:16Unidentified speaker (speaker_1): Thank you. Uh of course that's a variety of different things. That is yeah, um I'm doing Acorast. Acarast. Acorast Acker AST is a big one for me still, but I'm also mining while I

44:25Unidentified speaker (UNKNOWN): do

44:25Unidentified speaker (speaker_1): Acorast Ackerast because I don't do core. Um they probably would

44:29Unidentified speaker (UNKNOWN): like to.

44:33Unidentified speaker (speaker_1): Because if you

44:33Unidentified speaker (UNKNOWN): do

44:33Unidentified speaker (speaker_1): core, it like it locks down the phone and then you're limited to just core. Now, for some people, that's perfect, right? Do that if you want to sit and forget, put put it on core and just get started that way. Um I I have a little more time on my hands, so like I'm okay managing. uh these. I say that lightly. I I I have time to an extent, but um I can't so I have them running an Acoras in light mode, and then in the foreground I put AckerS, but in the background I have four to eight cores mining, Varius or Monero at any given point. Um so that doubles up on revenue, and then on top of that, you can pin a core on like one of the 20 phones to then run your server. So like the game server that's currently running makes around $500 a month recurring. Um, so that's all counted into that. That's Acaras, that's mining, um, all kinds of different things. Where sometimes you get one offs. Like a couple of years ago, um, Charles Hawkinson, the creator of Cardano, or one of them, um, there's a few, uh, he made kind of a push with this

45:37uh cryptocurrency called Nike. night, uh Midnight Network, it was.

45:42Unidentified speaker (UNKNOWN): Um,

45:42Unidentified speaker (speaker_1): and it was kind of like a gamified pre-mine, not a pre-mine, like a gamified mine right away. Um, where it was

45:50Unidentified speaker (UNKNOWN): uh

45:51Unidentified speaker (speaker_1): they allowed mining for two weeks. Where having the amount of phones that I have, where if you have a hundred phones, where you're not gonna have a hundred GPUs, I'm assuming for the average. home miner probably won't have a hundred GPUs uh sitting there ready to go. But if you have a hundred phones, sometimes mass device quantity beats out having you know an efficient, like really powerful device. Um because I was able to get 100 of them up um and I was kind of late. I was like six days in to the two uh two-week period, 14 days. Um, I was a little late and they kind of scaled it where if you were in the first day, you made more and then less, less, less, less all the way through. Um I I pulled in, gosh, around like $45,000 worth of night. Um, and with just and I was lazy, unfortunately. I I was being lazy that day, and I got up 200 phones and I couldn't. could have put on so many more. Um and I was like, oh, I'll just throw 200 on. Um

46:55Unidentified speaker (UNKNOWN): and

46:56Unidentified speaker (speaker_1): yeah, like over the course of the two weeks, it turned out to be like $45,000 worth of night token. Of course, then uh what a lot of cryptos do, and we we fall victim to is they have your lockup periods. So over time it looks great. And you know, on paper and hopefully depending how your tax uh taxes. is in your state and country um sometimes like you have to write that down as you know that was what I received I received forty five thousand dollars of initial cost and then by the time it's unlocked it's unlocked at like ten thousand dollars and you're sitting with a thirty five thousand dollar you know difference there um so it's you sometimes you have to be you know careful with what you get into and um to each their own. Like if you if those things come around and you have the devices, having the devices I would say has benefited me more than not having having the devices. Um because even if they're sitting there doing nothing, I think that's relatively fine for phones because what you had mentioned and obviously what we know is there's it's not just one thing, it's not an ASIC where you mine one thing. right? So the the usability, and I don't think I would do cell hasher personally or even sell these devices if it was only one thing because that's just your risk on that to it not being

48:14functional in a few years is just so great, sadly, because when it's one thing you're waiting for the next generation of model to come out for the one thing. And if that one thing isn't cooking, um like to to be fair, like if we're in if we're in a bull or excuse me, a bear market and uh your ASIC isn't cooking, you do have a device that's just sitting there, you know, doing nothing that is pretty expensive to say the least, um, all things considered. And then you when it does come to the bear or bull market, Mark, excuse me, there's probably another generation um right on the heels of that where um if you didn't sell your devices then to try to get the new generations, you might be caught in the pickup game of you know running your old devices, which is which is fine. I mean, people who have solar um and other people have ran old hardware for a long time and it works for them. And um yeah, I think leaving your hardware on is probably a better way than not leaving your hardware on, to be you know, fully honest there. But yeah.

49:13Unidentified speaker (UNKNOWN): yep

49:14Unidentified speaker (speaker_1): yep

49:14Unidentified speaker (speaker_0): so last question I have for you for today who are who is this world of phones for and who is it not for and who is it not for from your from your perspective?

49:25Unidentified speaker (speaker_1): Yeah I'm gonna I'm gonna say this very generously. If you think that you're the person who's going to have to message support every few weeks um and I like continuously You can totally do that, but it might give you a little more heartache than you would like. If you're familiar with GPUs, setting up GPU minings, if you're a Linux connoisseur, this is going to be right up your alley to an extent as well. It's still going to be a learning uh learning gap. Um There are easier ways to go. Like we do have products that are like to get you. started where you don't need to take apart the phones and you can kind of get introduced. So I would recommend then if you're going to even think about it, do that first because it's going to be cheaper on you if you don't want to do it. But it's also going to get it's going to be easier on you as well. So if that's going the way you would like um then definitely get the phones that don't have uh batteries and screens.

50:27and go that route and scale it at that. So yeah, if you're if you're a tinkerer or you you gotta if you're if you're stubborn to kind of work through things yourself and uh you want to figure it out, then yeah, go go for it. Um we do have it relatively as as easy as we possibly can, possibly can have it, but um every phone is different. So we've worked with a lot of phones, so yes, you can ask us questions. on that. But if you buy a DIY device and you don't get phones from us and you go and buy your own phones, we can help you to an extent, but you're gonna have to do a lot of the tinkering yourself. Because you're gonna have to find the operating voltage for it. And you know, if you get a phone that unfortunately can't operate within those voltage range, or it's one of the phones that does, but you have to power it on with the power a button. um after like a power outage or something like that, then that might be a little tedious for you. If you want to sit and forget, um probably not for you unless you're gonna do Acra S Core and you're gonna get just the basic charging station. But there are those options. So

51:30Unidentified speaker (speaker_0): yeah,

51:30Unidentified speaker (speaker_1): those are the

51:31Unidentified speaker (UNKNOWN): type of people.

51:31Unidentified speaker (speaker_1): Yeah.

51:32Unidentified speaker (speaker_0): That is right up my alley. I like the charging stations with the cell phones

51:36Unidentified speaker (UNKNOWN): that

51:36Unidentified speaker (speaker_0): solve the batteries in it. And that that's my work. world. And I call that lazy mining. It's like, listen, I don't want to go down the DIY route, I want to set it up, forget it, and I'm done with it. So it's good that you guys have different avenues for sure. There's definitely benefits of each. Um, but uh, yes, from my side of the world, I I absolutely love the charging station one. Well, listen, Luke, I appreciate you joining me today. Uh, thanks everybody. I'm gonna leave links directly down below. You can go check out Cellhasher. Cell Hasher. I also have multiple videos about Cell hashers, how to get started, what I'm using, the lazy basic versions. Uh, so go over and check those out, guys. And for myself and Luke, we'll see you guys next time. Take care.

52:15Unidentified speaker (speaker_2): That's going to wrap up this episode of the Everyday Miner podcast. If you enjoyed today's episode, please leave a review and a positive rating on whatever podcast platform you're listening on today. It goes a really long way. Finally, stop by the website thehobbyistminer.io for more crypto mining content, articles, videos, and much more. Happy mining.

This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights