No, this is not another article about Elizabeth Holmes

Nor is it a review of "The Dropout". Yet, given recent events, we had to put a little bit of thinking into the key issue behind this story: the "fake it 'til you make it" mantra that’s been driving a lot of early start-up successes over the past 20 years. Elizabeth Holmes followed this principle to the tee and yet it backfired pretty hard. Where did she go wrong? Are there any cases where this principle doesn't apply?

Very simply put, the "fake it 'til you make it" approach implies that there will be a time in the future where you’ll be able to make it. This is a great philosophy to abide by if your probability of succeeding is close to 100%. We all know some kind of start-up building a marketplace powered by AI. Except when you look closer, the marketplace is a nice landing page with a big spreadsheet in the backend that no one sees. And when you look even closer, the person behind the AI is a young, underpaid intern matching buyers and sellers in a pivot table before sending a manually automated email from "".

For most start-ups, this approach works really well. It shows grit, allows you to move fast and test lots of hypotheses without spending too much time developing an actual fully fledged product. That's why it's been used in Silicon Valley for so long. It somehow relates to the "do things that don't scale" approach so dear to Paul Graham and Y Combinator.

For other companies like Theranos, it doesn't.

Don't fake it 'til you know, like really know, like a 100% sure, you can make it.

These start-ups are what we're going to call, for lack of a better word, #deeptech start-ups. Deep tech start-ups are characterized by "the expressed objective of providing technology solutions based on substantial scientific or engineering challenges" (Wikipedia). This means the early days of your company aren’t about proving a market using pre-existing technologies, but instead creating a new technology that’ll heavily disrupt an existing market (e.g., the plant-selling business with Neoplants) or that will create a completely new market (e.g., the Augmented Reality business with Magic Leap, Neural-Interface Wearables with CTRL-labs or Wisear). When starting a deep tech company, there’s no way to know for sure if the scientific problem it addresses can successfully be cracked.

Sure, you can embark on a deep tech journey by bringing together the best team in the world, with the right knowledge, experience, and PhDs to give you the highest probability of success, but here's the sad truth: sometimes, you can’t overcome the physical rules of the world we live in (except if we all lived in a simulation, but this is something we’d be happy to discuss another time...). As start-up founders, early employees or investors, we need to be optimistic, otherwise we’d never start such a time and energy-consuming endeavor. We're naturally inclined to take any small sign as proof we're going in the right direction and we'll triumph over adversity:

  • A demo that works well in the lab on 70% of the users (though in the case of Theranos, the Edison machine of the early days didn't even reach that... Wait… Who said it's not an article about Elizabeth Holmes?)
  • A research paper proving a specific technique works for a specific population with 68% accuracy
  • A recent acquisition in the field from Meta, Google, Microsoft or Apple

Our optimistic brains tend to take a shortcut and logically conclude that, with the right amount of effort, we'll get to where we want to be. Unfortunately:

  • A demo that works in a specific controlled environment is far from being easy to generalize in the real world. There are so many parameters that can change, from interpersonal variability to environmental constraints you never even thought about. Until you've tested your tech in the real world and it's worked repeatedly, you can’t claim victory.
  • Research papers are a great way to make a first dent into a scientific challenge. They're meant to operate under a very constrained and measurable environment to be sure they can be easily peer-reviewed and the results can be reproduced. They definitely help pave the way towards future success of your deep tech start-up but, then again, until you manage to take it out of the lab and make it work on all your target users with a success rate close to 100%, you can’t assume you've won.
  • Market signs like big acquisitions are always a good sign, as they show interest in the technical problem you're trying to solve. Of course, the more people there are working on a specific topic, the more likely it is that someone will succeed in solving the technical challenge you've been investigating, but until they do... they don't.

With great knowledge comes great responsibility.

As a deep tech start-up, you HAVE TO be even more cautious with your claims. If you think about it, it's actually much easier for you to "fake it 'til you make it".


Well, you, your team and your scientific advisors are the experts, and very few people truly understand the technology readiness level of what you're demonstrating.

Your users, investors and the media usually don't have the technical know-how to understand if what you're showing is something that works for all your target users in any environment, or if it's just something that’s constrained to the lab.

Truth is that your users, investors and the media SHOULDN'T HAVE to get to that level of know-how (imagine if they had to do a PhD to understand every deep tech company they're looking at...), because they should be able to believe every word you're saying, because you should always stick to the unsexy raw truth of where you're at and what scientific blockers need to be overcome. Cases like Theranos should remain anecdotal as they hurt the trust placed in deep tech start-ups and in all the great entrepreneurs and teams who are fighting to push the boundaries of science.

This transparency and science-first philosophy need to be reflected in your company values. Otherwise, how can you expect your employees to be honest and transparent about their work if you’re okay with lying to your stakeholders?

Final Words - An Ode To All Our Fellow DeepTech Entrepreneurs.

At Wisear, we've decided right from the start that we’d be completely transparent about the progress we're sharing in frequent newsletters to users and investors (click here to get informed). Building the future of human-computer interfaces has its own share of scientific challenges. Being able to read brains, eyes and muscles from a device as tiny as an earphone is no easy task.

Before reaching the fully-functional tech we can now proudly display, we faced our fair share of unsuccessful live demos because of electromagnetic noise from laptops, earphones, cameras, and Wi-Fi routers around us messing up the signal our neural-based earphones would capture (videos available on demand...). However, we carried on doing these demos because they’re our reality checks. We're not here to show off with a quick good-looking performance but to build a real product designed to make everyone's life easier by providing them with seamless, voiceless and touchless controls over their devices.

Fortunately, there are plenty of fellow start-ups and entrepreneurs out there who are working with the same ethos. We'd like to finish off by celebrating the work of Dr. Steven LeBoeuf from Valencell, Inc.,Lionel from Neoplants, Séverine & Simon from IDUN Technologies, Paul & Maxime from MOTEN Technologies, Jean-Louis from Wandercraft and Yassine from Kinetix.

#deeptech #startups #entrepreneur #neuralinterface #entrepreneurship

About Wisear:

Wisear is developing the future of human-computer interfaces with the first wireless earphones equipped with a neural interface, delivering high-speed, private and accessible controls to users.

Wisear technology makes life easier:

Below are demos showing how people can already get voiceless/touchless control over their earphones and AR/VR/XR headsets.

Wisear's innovative solution was recently recognized as one of the best new technologies in 2022 by the Wall Street Journal, Techcrunch, Mint, The Telegraph, 20minutes and Sifted.