Shoutout to ALL projects building real tech / deep tech
Shared by balancegaming.network · 28d ago · 15 comments

Im writing this post to say that I appreciate everyone working on serious problems where it's difficult to make significant progress on a weekly or even monthly basis i.e. complex machine learning models, anything considered properly deep tech or applied research, hardware and so on where it will take forever to even get to the SaaS style metrics everyone expects. I try to vote you guys up whenever I can.

The Pioneer format puts you guys at a disadvantage and in another life I might have been one of you guys too (which is why I can often give at least marginally useful feedback lol). We totally shouldnt be ranked over a lot of you guys right now but it's just the game.

I made my particular startup choice because I recognize that Silicon Valley has become Metrics Valley now and tech has become a liability. You can get funded with metrics and no tech but you cant get funded with tech and no metrics. Feature companies, CRUD apps and stuff stitched together via API's with other CRUD apps reign supreme. Build metrics. Not technology

ptmn · 1d ago

Startup is a surviving game with extremely constraint resources. The goal is fighting against the odd by taking high-stake risk.

And tech isn't everything though people in the Valley speak in that language. Especially in the case of hard tech, the claims need hard evidences to prove (not social proof).

Forget what you read on news and media about fancy tech on the internet. Most of them are false perception.

The lessons I learn from my own experience and cross-check with several startup & tech communities (including top world-class grad schools with strong core team of deep tech) by real people with real stories are the opposite. Yes, metrics do matter, but they are different from business models. Know yours!

Put users and customers first. Spend time with them, talk to them, learn from them and join their daily life, and eventually you shall find the path.

ninogancitano · 26d ago

Great point about certain projects being at a huge disadvantage due to their longer development cycles!

I have found success on Pioneer by basing my submissions around my Asana projects. On Asana, I have a big-picture long-term Goals & Milestones board that take on average 3-6 months to complete. Then, inside those larger projects, are 1-2 week sprints that give me clear goals to aim for every week! Inside those sprints are 30-min to 4-hour tasks that I can focus on one-at-a-time on a daily scale! Can literally be something as simple as "install Python environment on new Windows computer".

A handful of those smaller tasks are what should go in your Pioneer updates! Something people can read and understand as well as something you can clearly define as "Done" or "In progress" or "Failed". If something is in-progress for more than 2 weeks, it belongs in Goals & Milestones, not Sprints! If it takes longer than half-day, break it down into simpler tasks until it does!

thanks for the laugh for "install python on new windows computer" :D That takes longer than to write than to do.. usually XD Good way to break things down and food for thought. I probably need to get more granular myself.

Chris_Wray · 25d ago

What is real tech? I think that good start ups solve problems or help people. Just because they are simple or not "hard" doesn't mean they aren't good, so have to disagree with you here. Now, if a person is building something that is hard, but can solve a huge problem, yes they deserve to be a pioneer.

Real tech to me is also something that is defensible because it was so hard/time consuming and reverse engineering it would take similar effort to just engineering it from scratch. Yes should solve a real problem or be very difficult but massive breakthrough, that's where the true value is. Not inflated shit like snapchat and pinterest

aaronggreen · 22d ago

I disagree with you that you cant get funded with tech and no metrics. For the right tech solving the right problems you will find investors. Look at openwater.cc (one of my personal favorite deep tech co's). Their dev timeline is probably somewhere in the 4-5 year range before commercialization and they have a pile of investors on board and more looking to join. I do agree that this platform favors quick, demonstrable progress, but perhaps there is a way to show progress on your deep tech that is engaging and demonstrable? I'm just spitballing here but maybe a few new lines of code you can share or rendering or something like that. I've worked in med device robotics in the past and I know how much effort and progress that can be!

its possible but deep tech funds are outnumbered by all the other comparatively dumb money 10 to 1 :)

andrewbirnbaum · 22d ago

Well said, it's interesting the point you make about progress. We have basically finished our framework (6+ months and over $500k) and have a decent MVP. Granted we still have a ton of CI/CD, and probably always will. We are being classified by so many purely based on "pre-revenue", versus the stage we are at, what we've done to this point, the team we have assembled, etc. We also wrote our own code and integrated blockchain into the platform. If Silicon Valley Vulture Capitalists can't see beyond certain metrics, their loss. There is so much great tech out there that really solves problems, I'm going to look to the solution and congratulate them on their successes, rather than all the negative that's out there, of which there is plenty! Thank you again for your comment, well said.

Ive recommended this to some of you in feedback but these guys just broke news of raising an even bigger fund. Many of you would be a better fit for Engineering Capital than Pioneer. I love that they even exist. Yeah my startup is designed for all the other dumb money metrics types and for people like a former client that said to me "I can do ads bro" when 2 months of coding was too much for him to keep rolling money into haha. https://techcrunch.com/2020/07/08/this-vc-just-closed-on-60-million-to-fund-technical-risk-saying-other-vcs-rarely-do-the-same/

hugobesley · 27d ago

Well said.

A neat solution to this natural, and unfortunate bias, would be to publish technical metrics. For example, an ML project could report the number of regressions they ran, or the change in their confidence intervals, or the number of rows of data they cleaned. I would be equally impressed by someone cleaned millions of rows of data as getting their first revenue :)

It depends is running regressions and tweaking confidence is really making them any progress or they're just pissing around essentially hoping to see some incremental result overall. I would say that row cleaning is more guaranteed progress than any of the other two hahaha

gavinyue · 27d ago

For me, my project deeplint.com is an open-source tool which makes it very difficult to input the progress since I am under development so far.

But also get some good feedback from time to time. Especially, when talking about marketing and sales, I guess the mythology is kinda similar.

Touchgram_Andy · 28d ago

Curious as to whether you regard Touchgram as deep tech or not? As a cross between Flash and Hypercard driven by gestures, it's fairly challenging programming at times. You could argue anything one guy can build is not deep :-/

My scale of difficult/deep tech ranges from glorified landing pages built with Bubble through to people modeling the planet in FORTRAN on supercomputers - I'd put my current efforts at the shallower end of that.

Its definitely not a CRUD app at least so at least you're doing something besides angular/react connected to Mongo like most people lol. Yeah yours takes more work. I think I remember questioning how useful it was and how many people had adopted/found it useful. One guy can build something deep but it might take a long time, deepness is the idea's difficulty/degree of execution, not how many working on it

I.e. for us, Ive deliberately done a low technical risk thing but the # of people in our community + those that have paid money to initially validate the idea as something worth paying for are over 250 already and could be more if we focus more on growth right now.

brainless · 20d ago

This is the primary reason why I dropped out of the Pioneer game from this week. One interesting observation I want to add: There are two kinds of feedback I kept getting -

1. Majority (like 90%) feedback was that my progress had not enough numbers or that I am talking too much about tech, that my site is not nice enough
2. A really small number of people straight saw the issue I am trying to solve and expressed interest in trying out my product when it is MVP ready

You can perhaps imagine who are my customers and why the rest of the casual feedback does NOT matter a single penny. It is a loss of time for me as a solo founder, so I showed myself out. Since my product is B2B, I am looking at other channels to network with founders of growing startups.