Perez-Breva challenges simplistic advice to focus on users when a product category and its behavior do not yet exist. Users, entrepreneurs, and investors cannot reliably predict life five years ahead, so someone must hypothesize and construct that future while remaining honest about the assumptions involved.

A genuinely new product may have no established user.
User research cannot make respondents predict an unfamiliar future.
Invented assumptions are acceptable when they remain visible.
Constructing the future differs from claiming certainty about it.

“if what you're doing is that new, is there already a user for it? If not, you're just making it up.”
Luis Perez-Breva · 15:50

“Someone has to make up that future.”
Luis Perez-Breva · 17:13

#user research#future#assumptions#product innovation

Myth Buster44:26

If It Does Not Help Humanity, Should It Count as Technology?

Drawing on MIT’s founding conception, Perez-Breva defines technology as the use of science and knowledge to help humankind extend its reach and gain control over nature. He argues that society’s task is to make powerful tools playable and beneficial for more people rather than treating technological development as intrinsically valuable.

Technology should apply knowledge for human benefit.
Access and usability determine who can benefit from technical progress.
Powerful tools still require deliberate social scaling.
Current AI and big-data systems do not automatically produce feared economic outcomes.

“if it is not helping us, it's not technology at least according to this definition.”
Luis Perez-Breva · 45:19

“more people can play with it and benefit from it.”
Luis Perez-Breva · 45:32

#technology#human benefit#access#automation

Myth Buster1:08:25

A Formula for Innovation Can Only Reproduce Yesterday’s Cake

Perez-Breva sees a paradox in growing demand for innovation formulas: a recipe reliably reproduces something already known. Although professional life has become more siloed, abundant online information also lets project-driven learners cross those boundaries and construct their own education.

Formulaic processes tend to reproduce existing outcomes.
Students increasingly request recipes for inherently novel work.
Professional and academic silos create communication inefficiencies.
Online information supports project-led, cross-domain education.
A concrete problem can guide what someone needs to learn.

“if there is a formula, there is a recipe, it's then you're just going to bake the same cake someone day.”
Luis Perez-Breva · 1:09:20

“more and more people I believe can actually start to learn on their own and ignore recipes or said models”
Luis Perez-Breva · 1:09:40

#formulas#education#silos#self-directed learning

Hot Take· 3

Hot Take30:42

MIT’s Student Mindset Changes Radically Every Two Years

Perez-Breva says the assumptions and communication patterns of incoming students change much faster than conventional generational labels suggest. He responds by rebuilding the class from a blank slate every two years instead of teaching to an outdated picture of the audience.

Student mindsets can shift on a two-year cycle.
Technology may contribute, but no single explanation is sufficient.
Teachers must repeatedly discard assumptions about learners.
Course redesign improves meaningful communication.

“Student body changes every two years.”
Luis Perez-Breva · 31:20

“every two years you need to entirely clean the class and redo it from the grounds aps”
Luis Perez-Breva · 32:09

#education#students#teaching#generational change

Hot Take45:58

AI Needs a Jarvis Objective, Not Another Human-Imitation Test

Perez-Breva argues that classic markers of artificial general intelligence lost their usefulness when machine learning mastered games once considered decisive. He proposes a new objective: systems that augment human problem solving as Jarvis helps Tony Stark build and act, rather than intelligence pursued as an end in itself.

Game-playing performance did not require general intelligence.
Older definitions no longer provide a sufficient research objective.
Useful AI should help people accomplish previously inaccessible tasks.
Search and recommendation systems already offer limited forms of augmentation.

“We need a new objective because the way we define that generalized started difficial intelligence, which is the touring test and these problems problems is…”
Luis Perez-Breva · 46:52

“the objective is AI that helps us the way Jarvis helps Tony Stark become a superhero iron Man.”
Luis Perez-Breva · 47:10

#artificial intelligence#jarvis#machine learning#augmentation

Hot Take55:31

A Social Network’s Existing Position May Be Its Only Moat

Perez-Breva questions the durable competitive advantage of purely social platforms, which could be displaced by something fundamentally different. He sees Facebook and Google primarily as advertising companies and suggests the web may have recreated the banner-dominated environment it once claimed to replace.

Network position may be stronger than the underlying product advantage.
A fundamentally different social product could disrupt an incumbent.
Facebook and Google remain heavily dependent on advertising.
Behavioral targeting often advertises products customers already bought.
Current internet advertising resembles the banner-heavy 1990s.

“I have a hard time understanding, ah why they have any competitive advantage, also whatsoever acceptive where they are already”
Luis Perez-Breva · 56:00

“Now look at the Internet today, it's exactly the same as it was”
Luis Perez-Breva · 56:59

#facebook#social networks#advertising#competitive advantage

Explainer· 3

Explainer03:17

Why Impossible Projects Often Reflect Disciplinary Silos

Perez-Breva argues that apparently impossible projects are often combinations that established disciplines cannot yet assemble. Henry Ford’s objective was not merely to manufacture cars but to make them affordable, with production scale, finance, and wages serving that problem.

Siloed thinking makes feasible combinations look impossible.
Transformative innovations initially appear preposterous.
A clear human objective should govern technical and organizational means.
Affordability, not manufacturing alone, was Ford’s central problem.

“But everything you love about the innovation started out looking impossible.”
Luis Perez-Breva · 03:31

“Everything else, the scale, manufacturing, finance, increasing wages, was a means to an end, which is make cars affordable.”
Luis Perez-Breva · 04:46

#impossible projects#silos#henry ford#problem solving

Explainer07:28

Innovation Is a Label Society Applies After the Fact

Innovation is described as the eventual social certification of work that once looked wrong or preposterous. Bill Gates’s computer on every desk and in every home appeared implausible before computers became ubiquitous and economically transformative.

Innovation is the end result, not the initial activity.
Novel work is commonly dismissed before adoption.
Widespread normalization can make radical ideas seem obvious in hindsight.
Economic effects become visible long after the original proposition.

“So innovation is how people certify what you did.”
Luis Perez-Breva · 07:50

“The idea of a computer and every desk and every home seems old all of a sudden”
Luis Perez-Breva · 08:56

#innovation#adoption#bill gates#hindsight

Explainer18:48

MIT Students Search for a Technology’s Hidden Superpower

In MIT’s Innovation Teams program, students from science, engineering, business, and architecture begin with a real laboratory technology. They determine the capability hidden inside it and search for a problem where that capability can matter, producing an actionable path that may differ from what the researchers imagined.

A laboratory discovery proves that something new is possible.
Its most useful capability and target problem may both be unknown.
Cross-disciplinary teams search for the capability-problem match.
The output is a path to modify and apply the technology.
Startups may emerge, but education is the program’s primary objective.

“somewhere they're hidden is a superpower. We just don't know what the superpower is.”
Luis Perez-Breva · 19:59

“that superpower is going to solve a problem somewhere. We just don't know where nor what the problem is.”
Luis Perez-Breva · 20:04

#mit#technology transfer#problem discovery#education

Story· 2

Story24:33

Nine Months of Emails Turned a Challenge into an MIT Opportunity

While finishing a master’s degree in Spain, Perez-Breva repeatedly contacted an MIT professor working across artificial intelligence and mathematics. He treated visa and logistical objections as challenges rather than rejections, ultimately forcing a clear decision and receiving permission to visit MIT.

He pursued work outside his formal chemical-engineering background.
The exchange continued for nine months.
He distinguished practical obstacles from an explicit rejection.
The visit began a career of moving between disciplines and MIT.

“none of them was no. Everything was a challenge.”
Luis Perez-Breva · 25:10

“what your study should not constrain you”
Luis Perez-Breva · 26:19

#career#persistence#mit#interdisciplinary

Story1:01:43

How Rewriting Mastermind Taught Perez-Breva to Program

At eight, Perez-Breva repeatedly retyped a Mastermind game because the family computer could not save it. He modified the program so he could always win and learned to code through the desire to play, shaping his later belief that complex technology becomes approachable when made playable.

His mother first entered the game program for him.
The computer could not preserve the code between sessions.
Repeated copying led to modification and understanding.
Play became his route into programming and AI.
Making complicated systems playable changes how people approach them.

“before you know it, I actually knew how to cod it. Basically, I was eight years old, mostly because I just wanted to play um…”
Luis Perez-Breva · 1:02:20

“The ability to play make things that seem complicated playable changes how we actually approached them.”
Luis Perez-Breva · 1:02:37

#programming#play#childhood#mastermind

Takeaway· 2

Takeaway23:00

The Lone Inventor Must Eventually Learn to Work Across Disciplines

A laboratory inventor does not automatically become an innovator. Bringing a discovery into the world requires learning outside the lab, changing one’s use of a discipline, and connecting with people who can help create an organization.

Invention and innovation are not synonymous.
Outside perspectives reveal knowledge unavailable in the lab.
Innovators must be willing to change their original conception.
Organizations require interaction across disciplines.
Team experience reveals both collaboration strengths and incompatibilities.

“They have to be ready to change the idea.”
Luis Perez-Breva · 23:36

“you need to learn how to connect with other people”
Luis Perez-Breva · 24:00

#teams#collaboration#inventors#disciplines

Takeaway49:48

Employee Innovation Time Should Produce a Resource Decision

Unstructured time for employees to present ideas may be inefficient. Perez-Breva says management should instead ask workers to explain how resources would prove a proposition at the next scale, converting creative time into evidence-oriented organizational development.

Idea presentations alone are weak outputs for innovation programs.
Management should request a path to the next scale.
Resource allocation should be tied to a concrete proof objective.
Prepared and motivated employees can innovate continuously.

“if they are just presenting ideas, I don't think it's efficient.”
Luis Perez-Breva · 50:26

“can you please tell me how we will direct resources to prove this thing that you have at the next scale.”
Luis Perez-Breva · 50:38

#management#workforce#innovation programs#resource allocation

Frameworks in this episode

Failure-Catching Scale-Up

Hunch-to-Scale Innovation Process

Investor Zero Exploration Rule

Productively Wrong Learning Loop

Insights & moments

⚡Myth Buster· 3

Why Future Users Cannot Tell Innovators What to Build