Author: Titus

Categories
Artificial Intelligence

My take on ChatGPT and other LLMs Replacing Jobs

I’ve been working with data for the past year and a half, as I’m building my own company, so I can’t ignore the noise around the launch of OpenAI’s ChatGPT, and the fears it has spurred.

Hype

That’s the first word that comes to mind when I hear about Chat GPT, GPT-3, other large language models. Journalists and bloggers love reporting about disaster scenarios where AI takes over the world and puts all people out of jobs.

Leaving this luddite assertion aside, there have been a lot of conversations about how writers, marketers, PR folks, coders and other creative jobs are at risk from these language models. I can’t think of something more wrong than that. A truly creative work can’t be fully replaced (yet!) with any of these models, because, well, the creativity part.

Let me explain.

A language model works off existing patterns and biases, so it can’t stray too far off from existing work. Even if the stuff it creates looks new, it actually is an extension of words/images/strings seen in the past.

If you try to have the model write about topics that require deep expertise or are controversial, then you will be frustrated with the result. It will put out mostly platitudes, if you’re lucky, or just purely wrong advice.

But if a job requires the person to gather symptoms, fixed inputs and other clear instructions, with a similarly clear outcome, then it’s much easier for a model to execute the same task with easier/better outcomes.

A few examples of jobs at risk of automation, and where a model could provide a better experience:

  • Primary care doctor that gathers your symptoms and recommends a quick solution to try out or route to a specialist for further investigation
  • L1 customer care specialist that responds to simple inquiries and fixes quick refunds, or routes to the right higher level specialist, as needed
  • L1 IT support / debugging based on error messages
  • Inbound sales specialist asking qualifying questions for a prospect, given a well-known product use case list
  • Content moderators
  • Bookkeeping and most of accounting tasks, as well as doing taxes for common situations
  • Business analysts / data analysts that pull reports based on manager asks from existing available data
  • SEO strategist choosing keywords to focus on in well-known/saturated markets

Here are a few more that I was surprised to see, but made sense after reading the reasoning behind them: sound engineer, animators, support engineers.

But we’re not quite there yet. For now, GPT-3 and other LLMs are just writing aids.

Categories
Business Models

JoinSave.com – a risk review

I’m going to make money if you sign up using this link. If not, then this article just helps a cool business move forward. This is also not investment advice, do your own research, always.

A few months ago, I saw an ad about a new way to leverage savings accounts and get more out of my money, without taking on additional risk on the deposited amount.

Save, which is still onboarding early adopters, keeps your money in FDIC insured accounts and only invests the equivalent of your earned interest, which, at this point, is around 3.5% per year on the most generous savings accounts. (2023 update, since savings rates change with the FED rate).

This way, your deposits are safe, and you can’t lose them, while they invest the interest equivalent on the stock market, in an attempt to 2-3x the return on those funds.

For example, on $100,000 in a Discover Bank savings account, you’re making ~$290 right now, every month. Instead of depositing that interest into your account, Save would invest an equivalent amount in the stock market on your behalf and hope to achieve a return of $600-$700/month.

Now, unless the stock market goes to zero or drops and never goes back up in a year, it’s my view that I’m set to lose, when compared to my current Discover Bank situation:

  • $261/month @ 90% meltdown, annualized; so only making $29 / month in interest
  • $145/month @ 50% meltdown, annualized; so only making $145 / month in interest
  • $72.5/month @ 25% meltdown, annualized; so only making $217.5 / month in interest
  • $14.5/month @ 5% drop, annualized; so only making $275.5 / month in interest

…but never more than I deposited.

If, on the other hand, the market goes up, and I deposited my $100,000 in the Save account, I would add the following gains over one year, on top of the $290 I would have got from Discover:

  • $14.5/month @ 5% increase, annualized, so making $304.5/ month in interest
  • $29/month @ 10% increase, annualized, so making $319 / month in interest
  • $30.2/month @ 25% increase, annualized, so making $362.5 / month in interest

There are some tax advantages to this method, too. While traditional savings accounts that deposit interest every month are taxed as ordinary income, the returns I would get from a Save account would be considered capital gains. Ask your CPA what this means for you and your tax returns.

If you sign up through the referral program, we each receive a bonus of $5,000 of equivalent portfolio investments. This means your account gets credited for the portfolio returns on an extra $5,000 over a period of one year. Neat.

For me, this is a low risk, high reward potential activity, and it makes sense to save with Save.

I’ve seen a lot of people not understanding how this program works, so I hope this article clears it up a bit.

Photo by Annie Spratt on Unsplash

Edit:

A lot of people have emailed me to ask about some details to confirm the referral. I case you need this information, here it is:

Name: Titus Capilnean

Email: titus.capilnean (at) gmail.com

Categories
Digital Tips Finance

Avail.com – collect rent via ACH without a headache

TL;DR: We became landlords (again) shortly before leaving the US to nomad around South America for 8+ months in 2022. I found Avail, by Realtor.com to process ACH rent payments. It’s free for the tenant and $5/mo/unit for the landlord. If you use my referral link to sign up, we both get 10 months free.

Before my wife and I left the United States to travel as nomads around South America, we had to figure out what to do with our house in Austin. We didn’t want to sell it, but we also didn’t want it to sit empty while we were gone. That’s when we decided to become landlords and rent it out.

We had it on the market for a few weeks, first furnished, then unfurnished, until we found a family to enjoy our place as much as we did.

One of the things we had to do was find a way to handle the rent payments. We didn’t want to deal with checks, as they can be a hassle to work with, especially as we are traveling. Instead, we decided to look for a way to process ACH (Automated Clearing House) rent payments.

After some research, we found an early stage startup that seemed like it would be a good fit for us. They handled the ACH payments and made it easy for us to keep track of everything from afar. However, unfortunately, this startup folded after a few months, and we had to find another solution.

That’s when we discovered Avail. We tested their services for a few months, and we were really impressed with how easy they made it for us to manage the rent payments and communicate with our tenants. It’s free for the tenant to use, and it only costs $5 per month per unit for the landlord.

We’ve been using Avail since September, and it has been a lifesaver. It has made it easy for us to keep track of the rent payments and communicate with our tenants, even though we’re not in the country.

Overall, becoming landlords has been a great experience for us and using Avail has taken the guesswork out of the rental payment process.

P.S. the photo is of the actual house.

Categories
Digital Tips

Mint Mobile – A US number for digital nomads

I wrote this review from Argentina, after 8+ months of using Mint Mobile while traveling through South America.

I will get paid if you sign up via one of the links in this article, and so will you. Switch to Mint Mobile already and stop spending so much on your wireless plan. Use my referral link to get $15 in renewal credit.

Mint Mobile is a prepaid wireless carrier that has gained popularity among digital nomads for its affordable pricing and flexibility. One of the key features that makes Mint Mobile appealing to travelers like us is its support for WiFi calling and texting.

With WiFi calling and texting, you can make and receive calls and messages over a WiFi connection, rather than relying on a traditional cellular connection. This can be especially useful for travelers that don’t want to lose access to their main phone number.

You can swap SIM cards between Mint and the local network you get on as you travel to new countries, or you can use a second phone as a hotspot for your Mint one if you are not near a WiFi.

To use WiFi calling and texting with Mint Mobile and not have to remove your SIM card every time, you’ll need a DSDS (Dual SIM Dual Standby) or DSDA (Dual SIM Dual Active) enabled phone. These types of phones allow you to use two SIM cards at the same time, one virtual eSIM, one physical SIM. It’s fairly easy and quick to install a Mint Mobile eSIM on your phone. It took 5 minutes on my Google Pixel 5.

This means you can have one eSIM for your Mint Mobile service and another for a local physical SIM card when you are traveling abroad. This can help you save on roaming charges and stay connected when you are outside of the United States.

The main benefit for us is that we can keep your two-factor authentication enabled. Two-factor authentication is an extra layer of security that requires you to enter a code sent to your phone in addition to your login credentials – like for banks and credit score accounts. This can help protect your accounts from unauthorized access, but it can be inconvenient if you don’t have access to your phone. With WiFi calling and texting, you can still receive the codes you need to log in to your accounts, even if you don’t have a cellular connection.

Overall, Mint Mobile is a great choice for digital nomads who want an affordable and flexible wireless plan that includes support for WiFi calling and texting. Whether you’re working from a coffee shop in Argentina or taking a call from a client in Colombia, Mint Mobile has you covered.

P.S. If you’re using a VPN service on your device, make sure to turn if off on WiFi at least until your WiFi calling and texting becomes available. Then turn the VPN back on.

Photo by Nestor Barbitta on Unsplash

Categories
Venture Capital

Investing criteria: Can this company produce a +$1B exit?

This year, one of the things that I managed to do was to structure the way I evaluate a company – for investment, partnership or even as a potential employer. This investing criteria framework is a quick way to decide if it’s worth putting money in a company or not, based on a $1B exit goal.

The Investment Criteria

  • Strong Founding Team
  • Team diversity and size
  • Traction / Interest
  • Market size and size of pain
  • Investor signaling, valuation velocity
  • Timing

The Example Startups

  • Mux – video streaming API tool
  • PolicyGenius – insure-tech comparison marketplace
  • CrowdAI – data preparation for computer vision
  • Chatdesk – eCommerce customer support tool

Strong founding team

I look for signaling here – accelerators, incubators, good schools and diverse backgrounds. This de-risks some of you selection bias, since they have been through filtering before. You want to make sure the team is complementary enough and that they get along well. Serial entrepreneurs with successful exits are great, too, since they have provided investor returns before. I also look for relevant go-to-market experience and in-market expertise, when looking at founding teams, to ensure they are able to execute not just the product part, but the distribution.

Mux: YC, Brightcove alums, serial entrepreneur, team worked together before. Strong industry experience

PolicyGenius: Diverse founding team, ex-McKinsey, worked together previously, ex-Harvard educated

CrowdAI: Two people ex-Google, one of the founders had a minor exit in AdTech.

Chatdesk: ex-Google, McKinsey, Vanguard founders, but lacking go to market key people

Team diversity and size

This first investing criteria is important regardless of the size. More diverse teams produce better results in the long run, so going beyond the white male group is important for my investment criteria. I look at both the founding team and their early employees. LinkedIn is a good way to empirically test this, but Crunchbase has started adding diversity signals too.

Mux: Diverse team, but only white male founders. 61 team members to date, gender diversity in the engineering team. Good industry experience for most team members

PolicyGenius: Gender diverse teams, large organization, focus on people and operations, 395 people and growing

CrowdAI: Diverse founding team – gender, ethnicity; small team (20), mostly male; not significant industry experience i.e. strategy person coming from oil and gas

Chatdesk: Diverse founders and team – gender and ethnicity, global operations, strong data science backgrounds. Small team (26). Very little Go-to-market staff headcount and lack of industry experience

Traction / Interest

This is where the evaluation depends stage-by-stage, company-by-company and industry-by-industry. If there are visible signs of paying customers, like logos on the website, and notable case studies, then that’s always a great sign. I also try to look at their web traffic, trends, industry keywords where they show up via SEO, and signs of an active community – on Twitter, Discord, Github, LinkedIn, or any other platform. No activity, especially for community-dependent products, is a red flag.

Mux: Customers like Vimeo, TED, Udemy, Wistia, Reddit, Fox, Robinhood, Scale, Equinox, Amazon as channel partner; 140,000 visitors/month on their website, strong growth month over month, hot industry (API-based)

PolicyGenius: 1.2M visitors / month, over 50% US (core target market); new partnership announcements with the likes of Brighthouse Financial

CrowdAI: Low traffic, low time on site, high bounce rate, no marquee customers

Chatdesk: 13k/visitors per month, but mostly non-US, US traffic accounting for 20%. US growth in traffic recently (4x), a few niche customers in the ecommerce space on their website, as well as case studies, large number of integrations with existing support tools. AI powered

Market size / Size of the pain

Usually, if startups get this far down the decision making path, then they are on to something. But even if they have a great product, team and traction, the market might not be big enough for a $1B exit to happen. Or if the market is big, but the pain is not real – a nice to have vs a must have product, then they aren’t likely to grow into something huge. Competition is a good sign here, if they are able to differentiate and serve the pain in a more efficient, compelling or interesting way. If the product is actually a feature that most people in the target market can live without, then it’s safer to pass.

Mux: +$40B market size for streaming alone, and growing with COVID19 pushing people to rely on video for almost all interactions; Streaming is expensive and their pay-per-use system works for teams big and small

PolicyGenius: +$1 trillion market size in the US alone, hugely outdated and cartel-y; A few key competitors, like Lemonade, which recently went public. Large pain

CrowdAI: Large market (+$25B by 2025), big pain with computer vision being used more and more across all industries, and tech-assisted data labeling being a critical path to growth

Chatdesk: $73B market for customer service automation tools, however incredibly crowded and little differentiation among players. Relatively small impact doesn’t alleviate the customer service pain.

Investor signaling, valuation velocity

Again, this is stage dependent, and if you’re looking at this, then that means the previous hurdles were a pass. If I still have doubts about the market/traction/product/team, then this filter is pointless. When I look at investor signaling and valuation velocity (assuming there were a few funding rounds before the evaluation), I look for top tier VCs/Accelerators/Angels that put money into the company, are advisors. I also look for a stage-to-stage 2x valuation increase, or more. There are industries where valuations will lag early on and then as the company gains more traction, they shoot up. But if you see anything north of 2x round over round, then it’s a good sign.

Mux: Just closed a series C for $37M, with a total of $68.9M to date, with big names like A16Z, Accel joining Susa Ventures in later rounds. PE backing through a TPG Growth arm in Series B round; Valuation in the late tens of millions, early hundreds of millions.

PolicyGenius: 20 investors to date, $161M raised at triple digit millions valuation, including PE backing, signaling potential profitability, on route to unicorn status, strong traction for growth

CrowdAI: Only $2.1M raised to date in 4 years, small team, implied slow growth

Chatdesk: Only 2 funding rounds to date in 3 years, $2M raised, implied slow growth, small team

Timing Criteria

Last but not least, are they starting the company at the right time? Is there enough critical mass, infrastructure, capital, customer appetite to generate growth? This one is harder to quantify and it requires you, as an investor, to look at the market and try to understand its dynamics. Where is it on the Gartner Hype Cycle? Is there even analyst coverage for this sector? What signals can you read from key players?

Mux: Founded in 2016, they are positioned to grow with the COVID19 video tailwind, and have landed sufficient traction to attract even the most picky buyers. Their usage-based, AWS-style pricing model makes sense and they are well positioned to grow

PolicyGenius: Started ahead of the game in 2014, and now benefiting from insurance-tech adoption tailwind and positive COVID19 impact due to people moving away from brick and mortar agents

CrowdAI: Early to market in 2016, but didn’t scale fast enough, and are now being overtaken by competitors.

Chatdesk: Late to market, with a lot of customer service automation hype already captured by Salesforce Einstein and IBM Watson in 2016-2017.

Final verdict for Example Startups

Mux: High unicorn potential.

PolicyGenius: High unicorn potential.

CrowdAI: Low unicorn potential.

Chatdesk: Low unicorn potential.

So far, these investing criteria filters have proven useful for my angel investing and advisory engagements. I’m always looking to improve them, so if I missed any other criteria that you found useful, comment below. I’d love to listen to other perspectives on this.

Photo by De’Andre Bush on Unsplash

Categories
Politics

Tax Breaks, not Taxes for People who Work From Home #WFH

Recently, a Deutsche Bank analyst decided to put a note out there that governments should tax people that are able to work from home.

They suggested a 5% tax is due for the privilege of being able to work from home during the pandemic, and for not participating in the large economy that forms around workplaces – dining, coffee, public and private transportation, parking, gas etc.

For me, this type of analysis is disconnected from the reality in most people’s lives. Not everyone works for VC-funded tech startups or FAANG companies that offer generous stipends for home offices, so working from home has meant that expenses have gone up for most people. Sure, we don’t commute anymore, but we spend 100% of the time at home, which carries other issues.

Our WFH life

Here’s what working from home looks like for me and my wife:

  • we had to get a 2 BR apartment instead of a 1 BR to be able to work from home without interrupting each other, so extra money on rent
  • our two laptops and multiple monitors run all day long, so we use more electricity and had to update our internet connection to sustain 2x video calls
  • we spend 100% of the time at home, so we have to run the heating all day, too, since it’s autumn/winter now. We don’t have AC in California, but we had considered a portable one during the heatwave.
  • we spend more money on food, now that we have to buy and cook for the whole week, order out
  • we run the dishwasher daily, if not multiple times per day given that we eat 3 meals x2 every day at home

All these things translate into increase cost of living at home, more wear and tear to our homes, on top of the impact the lockdowns and social distancing has had on our lifestyles. We are thankful, though, to be able to work in industries that enable this kind of work.

Tax Brakes, not Taxes

But back to taxes or tax breaks. Instead of taking more money from everyone working from home, I propose governments do the following (we’re based in the US, so I’ll take the IRS as the example here, but these can apply to any democracy):

  • Enable W2 workers to take work from home deductions – like 1/3 of the rent, as business expense, since we’re using it as an office. For now, only 1099 workers or LLCs can do that.
  • If you’re a homeowner and have fully paid your place, great! You should also be able to deduct an extra depreciation due to working from home all day
  • We should be able to do the same for electricity, internet and items we bought for work – desk, chair, new monitor, keyboard, mouse etc.

This would free up capital for people to invest, save and better prepare for time where they might be unemployed or for some other life event.

Another tax would only make middle class people work more for their money, and provide extra funding for unclear government goals that might transfer wealth to ultra high net worth individuals.

So no, Deutsche Bank, we shouldn’t tax WFH more.

Photo by The New York Public Library on Unsplash

Categories
Life

How I network in the time of COVID-19

I was talking to a founder earlier today about networking in the time of COVID-19 and he suggested I write a post about it.

Before the pandemic hit, I was out there every week, attending events, meeting new people and networking all over the US. I even ran a tech-focused group in San Francisco for about 4 years.

When the virus hit, it all suddenly went away. At first, I didn’t think too much of it, continued to talk to people via social media, video calls and emails.

But as the situation continued, I started feeling the itch again, to meet new people, explore new ideas and just network. I tried to go to a few industry-specific meetups that were now happening virtually, but it just didn’t cut it. They were usually centered around a speaker, or got polarized around a common topic, or, worse, they got dominated by someone very loud, with poor listening skills.

So I dropped that, too.

I later discovered Lunchclub. They apparently started out as an in-person networking network (ha!), but COVID-19 had them pivot to online only. The way it works is:

  • you sign up, fill in your profile and interests, set your location
  • you select time slots that work for you. Make sure to block out the work-calendar, too
  • you get matched every week with someone new, based on your interest
  • you meet at the agreed-upon time. Or reschedule
  • you provide feedback about your match

They have a points system, with which you can have more choice when it comes to meeting certain people from certain areas. You can earn points by inviting other people to the network. Clever.

My experience with them, after 5 meetings, has been mostly positive. Three of the meetings have been extremely relevant, two so and so. But all have been interesting people with cool projects, life stories and ideas to discuss. I surely wasn’t bored on any of the calls, which is not what I can say about some networking events I used to attend in person!

It’s not the best time to be out there networking in-person, so this has been my approach to date. Looking forward to getting back out there without the fear of deadly infections arising from any larger event.

Photo by Antenna on Unsplash

Categories
Finance

Own the lottery with Yotta – a place where people can save and win

This is not investment advice, and as part of the Yotta account, both the new user and I earn 100 tickets with my referral link (or the code TITUS2) once the new user deposits at least $25, for 1 ticket.

A few weeks ago, I discovered a new type of savings account. Savings are usually boring, and I’m glad more and more fintechs are now focusing on making the FDIC insured account more cool. Compounding interest is great, but when it only yields 0.6% per year, it takes a looooong time to make real money. For example, if you deposit $100 in a high yield savings account, currently at 0.6%, and leave it there for 20 years, you’ll have $112. If the inflation rate is higher than 0.6%, then you’ll have lost money in real terms. Not the best deal.

I wrote about Save a few weeks ago, where they invest your interest to make higher returns in the stock market, while keeping the principal safe.

Pay and Win VS Save and Win

Now it’s time to talk about a different type of behavior – a negative risk-seeker, commonly known as the gambler. I’m talking about the people who buy lottery tickets as a regular habit. The ones who play the slot machines in Vegas. The ones who play backjack, roulette or poker with the hope of getting rich, while also getting those tasty dopamine hits every time there’s a game move.

Let’s focus just on the lotto players. The average American spends about $1,000 / year in lottery tickets that have an average win rate of 1 in several hundreds of millions of entries. They are more likely to get struck by lightning than to win anything at lotto. But they get their dopamine every time there’s a drawing, in exchange for that $80ish every month.

What if you could keep those $80 and still get the same dopamine hit?

I found Yotta, through a Youtube real-estate influencer, and I started looking into it. The startup is a YCombinator graduate, fresh on the market.

They combine an FDIC insured savings account with a guaranteed rate of 0.2% and a weekly lottery with prizes of up to $10M. For every $25 you deposit, you get a ticket with 7 numbers – 6 regular and a Yotta number. Every day, they reveal one number, at 6pm PT. Each Sunday afternoon, you can see how much you have won, all in the app. Simple and fun.

Two Phones
This is how their mobile app looks like.

The app easily connects to bank accounts through a service called Plaid, now part of Visa, so super-secure. While you can put in whatever you like, due to financial regulations, you can only deposit a maximum of $10,000 / day and can withdraw a maximum of $40,000 per month. And remember, FDIC only insures accounts of up to $250,000, so putting in more is riskier.

How does it work?

Technically, Yotta makes more on your money than the 0.2% that it’s committed to pay out on a regular basis. The rest of the profits they make from holding your money in risk-free assets they distribute through the lottery system – via actual cash or a Tesla Model 3.

If you calculate your odds of winning today, by playing consistently, and without too much network growth, you could be looking at yields of over 1.8% per year at worst, over 3% per year at best.

So, from my perspective, anyone who plays scratch cards, lottery tickets and any other form of small-value gambling should definitely try out the Yotta app and put that $80 per month money to work for them, instead of spending it.

Photo by dylan nolte on Unsplash

Categories
Blockchain

30 days of DeFi – Journaling my experiments with crypto and decentralized finance

If you haven’t experimented with these things yet, here’s a good place to get started with crypto. Also, this is not investment advice, you may lose all your money by trying these things out.

“Decentralized Finance (DeFi) is the merger of traditional bank services with decentralized technologies such as blockchain. DeFi can also go under the name Open Finance due to its inclusive format. Importantly, the DeFi community seeks to create alternatives to every financial service currently available. These services include items such as savings and checking accounts, loans, asset trading, insurance, and much more.”

According to Securities.io

Here’s also a more comprehensive guide to DeFi by Coinbase, if you want to learn more before reading.

It’s been over 30 days now since I started experimenting with decentralized finance tools on Ethereum. It’s been a wild ride and I tried to do a lot of things across the entire risk spectrum.

58 transactions and 0.82 ETH gas fees later, I can tell the story of how I won, lost, then won again, then barely edging out a little net win. Follow my story day by day.

I used Zapper.fi to explore all my DeFi experiments in one place.

August 2-3

This is when I decided to take the plunge. The market had just gone up in late July, and Twitter was full of chatter about yields to be made in decentralized finance tools (DeFi) like Compound (COMP), Curve, Aave and others. Up until then, I had done little research on any of it and had stayed away from MakerDAO and DAI.

After reading a lot on Twitter, I got onto InstaDapp, since it looked like the easiest way to run some experiments, thanks to their combined transaction model. Great UX, btw.

So that day I gathered all of my available ETH, and deposited it into an InstaDapp contract. I also deposited some BAT that I had earning peanuts in COMP, as collateral for safety. I’ll get to this later.

I borrowed DAI against some of the ETH, then I deposited that into Compound via InstaDapp.

August 4

I did some more research on ETH leveraging through InstaDapp and decided to wind my DAI position up by 15%, so I borrowed DAI, bought ETH and added it as collateral in my MakerDAO account. This will come to bite back later.

On Compound, I decided to try and 4x my mining earnings, so I leveraged my existing collateral (remember the extra BAT?) and borrowed USDC to get DAI and used DAI as collateral to earn COMP.

August 7

There’s a little dip in the market and my MakerDAO position turns from Safe to Risky. I have to get more ETH and add it to my collateral, because I’m already leveraged. No biggie, just some more gas burned. Let’s call it insurance.

I also bought some COMP off the market, that I would use later to farm. That’s when things get interesting.

August 10

Over the weekend, as Crypto.com releases its migration from MCO to CRO, I decide to use some of my BTC to get more collateral and upgrade my crypto card. I was already getting used to InstaDapp, so I get my BTC from my safe, move it to Coinlist, where I wrap it to wBTC, so I can add it to the ETH ecosystem.

With the wrapped BTC, I go on InstaDapp and get enough of a USDC loan to get CRO and update my card. The USDC loan at this point is @ 5% APR, but my COMP mining from it pays for that and more.

Cool. Free money, sans gas costs, which I pay.

August 11

I continue going down the Twitter DeFi rabbit hole and I discover Yam.Finance, a new monetary experiment with automatic rebasing currency, while farming liquidity tokens off Uniswap.exchange.

This farming works because liquidity providers get paid a fraction of the transaction costs, and on a very liquid market, you can make significant returns by just keeping coins in there as a market maker.

Back to Yam. I deposit COMP and start farming Yams. While I was getting ramped up with Yam, an error was found in the Yam smart contract, rendering it useless.

August 12

I exit the Yam farming contract, reap Yam rewards and attempt to delegate my yams to save the protocol. Gas prices soar, so my transaction is delayed and I miss the delegation snapshot. Expensive move. Also during this time, Yam rebases chaotically, driving huge price swings. I hold on.

Drama ensues in the Yam community and the team decides to migrate tokens to a new contract that doesn’t rebase.

August 19

After following the Yam groups, I decide to continue with them and move my Yams to Yam v2. More gas burned, but at this point it’s a sunk cost. I want to see where this goes, for the monetary experiment, if anything.

August 21

Guess what, the market dips again, so my InstaDapp risk rating drops very close to Risky again for the MakerDAO position. I add another bit of ETH as reserve in that smart contract, in case the price drops some more.

August 26

I randomly get some COMP off one of their promotions to learn about the protocol, and I get more insights into their governance structure and process. This will come handy later, too.

August 28

Someone decides to fork Uniswap, leverage the Yam UI (which was beautiful, btw) and the COMP governance model, and creates SushiSwap, which takes crypto Twitter by storm. Seeing traction and contract audits, I join the bandwagon.

Sushiswap essentially asks you to provide liquidity on Uniswap for a token set, like COMP-ETH, or SUSHI-ETH, and then take those liquidity tokens and deposit them. That deposit then earns you a 4 digit % yield in SUSHI. Their goal is to launch a community-led Uniswap, with profit sharing through the SUSHI token. Cool.

I set my target of 48 hours to harvest and cash out some rewards, having learned from the Yam debacle.

I also get some more ETH, since I’m burning through gas like crazy with all these smart contract executions.

August 31

Ah, the end of summer.

Also time to harvest those SUSHI. I provide more COMP-ETH liquidity on Uniswap, because I didn’t deposit everything on purpose on August 28. The Sushi contracts automatically harvest rewards when you deposit, kind of like Compound does it too.

I get my SUSHI, which I convert to ETH. I feel great about myself for the rest of the day.

Remember, at this point, I’m still farming SUSHI at 4 digit % returns, so I give it another 3 days this time before my next harvest.

At this point, there’s over $1B worth of crypto locked into SUSHI farms. What could go wrong?

September 3

All is well, bull market under way. As planned, I harvest more SUSHI and I add liquidity to Uniswap’s SUSHI-ETH pair. I then bring my LP tokens and start farming SUSHI with SUSHI.

At this point, there’s $1.5B in assets locked in SUSHI farming contracts. People are taking notice and everyone is excited for the swap launch, with community governance. Even Binance endorses it.

Also, the market is starting to slide again.

September 5

The market drops sharply after stocks take a nosedive at the end of the week, so my InstaDapp risk tool flashes red again for the MakerDAO position.

At the same time, the dev running Sushiswap pulls an exit scam move, by selling all his SUSHI, effectively crashing the price by 50%. I get my COMP-ETH out of there and remove liquidity, harvesting some SUSHI in the process.

Total assets held in Sushi contracts drops back to under $1B, by about 60%, too. I’m keeping my farming SUSHI-ETH pair, and some SUSHI free-float, in case I can take some more profits

I need the ETH for the MakerDAO position, so I add more ETH, which, combined with the previous reserve ETH, move me down to Safe again. Liquidation limit moves down at $244 / ETH.

Right now, when I finished writing this article, I’m up, if you consider the YAM and SUSHI I still hold, at current prices. But this changes every minute.

From a gas point of view, this was an expensive ride, but I learned a lot. And I barely scratched the surface. I will definitely try more things soon and report back.

Tools I found useful along the way

I already linked Zapper.fi at the top – great way to visualize assets and transactions.

I found this good article on ways to save on ETH gas price by timing your transactions. I also learned that if you transact with less decimals, you burn less gas.

Metamask – a Web3 wallet that I’ve been using since 2018. Even Bloomberg wrote about them recently. It’s my central command for ETH and ERC20 transactions.

Uniswap Exchange – fast-growing decentralized exchange for ERC20 tokens, and market making for normies.

Compound Finance – the OG place where I started playing with yield farming. Backed by A16Z, it’s one of the safest things you can do in crypto.

InstaDapp – my entry point into the leveraged DeFi world, they created recipes to maximize yield, wind up assets and profit from longs, liquidation risk notwithstanding.

wBTC network – a group of entities that enable people to wrap their bitcoin onto ERC20 tokens. Useful to get your bitcoin into DeFi.

MakerDAO – a place where you can deposit ETH and get DAI loans at 0% APR, which you can farm with. An interesting stablecoin that’s hugely popular.

DeFi Prime – content aggregator about the industry. Good to discover projects and experiments.

DeFi Pulse – check out the amount of money locked in DeFi contracts. Good to check project health.

DeFi Explore – a place where you can see your DAI exposure and profit/loss on the Maker CDP.

DeBank – a Zapper.fi alternative to viewing portfolios, making transactions. Has a cool Swap view of decentralized exchanges, with rates and gas fees.

Elrond – a promising chain to compete with ETH DeFi.

Zerion – visualize market stats. See how much is in each liquidity pool.

Zippo – another cool way to visualize projects, like SUSHI, YAM – price, yield, liquidity.

Projects I haven’t explored yet

Aave – didn’t really have attractive yields for me. Maybe I don’t know enough about it. It also only recently become available on InstaDapp.

Synthetix – too risky for my taste, I don’t understand enough about derivatives. SNX tools is also a good way to visualize that ecosystem.

Dia Data – open price feeds for DeFi

Idle.Finance – decentralized hedge fund for ERC20 & ETH, with risk scores.

Yearn Finance – a token that is more expensive than BTC now, and with 2 digit % yields on ETH, DAI.

Photo by Bermix Studio on Unsplash

Categories
Blockchain

Failed food deliveries – a technology fix

Today, a Caviar courier managed to not delivery my order. This is not the first time a courier fails to deliver my food, although it showed as confirmed in the app. I live on a little hilltop building, with an alley way entrance between two small lion statues. I’m aware it’s hard to find and I always add instructions on how to get to my apartment. It usually works, but sometimes I get to starve for another 30+ minutes until we figure out other alternatives to the meal we’d been waiting for.

In this day and age, with contactless deliveries, which are great, it’s hard for couriers to know that they have delivered to the right person. So we need a solution to enable them to verify the order destination without me being there.

Sounds like a job for zero-knowledge proofs, if you’re a crypto-geek, like me. Or it’s just a simple async identity verification problem.

There’s a problem with addresses. Your GPS might say you’re at one address, but you’re actually at another building. Some buildings have the numbers well-hidden, others don’t have any numbers altogether, so couriers have to guess.

Being in love with technology, and being an advisor for Tailpath, I propose this simple solution:

  • As a customer, I want to give the courier enough information for them to be able to verify my identity, without being face to face with them, but not have to print / show identifiers each time. I also don’t want to reveal my personal information at the door.
  • As a courier, I want an easy way to verify if I’m dropping off the order in the right location. But I don’t always know if that’s the right address. I can’t ask the customer due to COVID or because they are not home, so I need to rely on signage
  • As an app builder (i.e. Caviar, Uber Eats), I would like to have a reliable way for couriers and customers to make the transaction without physically meeting.
  • Customers could have their address on a QR code on their door, for couriers to scan with their Tailpath app, or a white-label version of it within Caviar or Uber Eats or similar. You’d only need to put the code up once, and reuse over and over again
  • If the address matches with the address given in the app, the delivery is confirmed and the courier knows they dropped off the order correctly, and the customer is notified. The courier can also take a photo of the delivery at the correct door, for proof.
  • If the address doesn’t match, the courier can call the customer for further instructions.

This way, the courier proves they delivered in the right place, and the customer has no way of arguing that there was a mis-delivery if the courier scanned the code at their door and left a photo proof of the delivery.

I hope these apps get better, so less people have to go through the hassle of not having their food delivered correctly.

Photo by ivan Torres on Unsplash