AI business ideas funded by YCombinator
Summary
TLDRIn this video, the speaker discusses how to determine what to build with AI, emphasizing the importance of natural aptitude, deep interest, and the potential for great success. Drawing insights from Y Combinator's portfolio, they analyze emerging trends in AI startups, focusing on industries like healthcare, fintech, and education. The video highlights the dominance of B2B solutions, the significance of technical expertise, and the untapped potential in areas like AI safety, data privacy, and blockchain integration. The speaker encourages leveraging generative AI innovatively and addressing ethical concerns to stand out in the competitive AI landscape.
Takeaways
- 😀 Paul Graham suggests that successful work with AI should align with natural aptitude, deep interest, and the potential for great work.
- 🔍 The speaker recommends observing successful AI companies backed by Y Combinator to identify emerging trends and technologies in AI.
- 📈 Healthcare and biotech are leading industries for AI startups, with 45 companies in the dataset, indicating a strong adoption of AI in these sectors.
- 🏢 A significant majority of AI startups (81.1%) are focused on B2B solutions, reflecting investor confidence in this sector.
- 🛠️ Infrastructure-focused AI startups are less common, with only 14.9% of the companies operating at this layer, suggesting higher barriers to entry.
- 🤖 Automation is the primary use case for AI across industries, with 69.1% of companies aiming to assist human work, while 31% are developing fully AI-driven solutions.
- 📊 Generative AI is the most popular technology among startups, with 78 companies leveraging it, likely due to advancements in large language models.
- 🔒 Addressing data privacy and security is an emerging area with potential, as only a small fraction of startups are currently focusing on these issues.
- 💡 Startups working on ethical AI or AI safety are rare, indicating a significant opportunity for innovation in these critical areas.
- 🌐 The integration of AI with emerging technologies like blockchain and quantum computing is still in its infancy, with only a handful of companies exploring these intersections.
- 🎓 The typical AI startup founder has strong technical expertise, often in computer science or related fields, and many have backgrounds in leading tech companies or academic research.
Q & A
What are the three qualities Paul Graham suggests a project should have according to the video?
-Paul Graham suggests that a project should have three qualities: 1) It should be something you have a natural aptitude for, 2) It must be something you have a deep interest in, and 3) It should offer scope to do great work.
What is the significance of looking at Y Combinator's portfolio to identify AI trends?
-Y Combinator's selection process has consistently surfaced companies that go on to reshape entire sectors, making their portfolio a valuable indicator of emerging trends and technologies in AI.
What does the video suggest about the current state of AI in terms of ideas and noise?
-The video suggests that AI is currently filled with noise due to hype and many bad ideas. It emphasizes the importance of finding good ideas amidst this noise.
How many AI companies were analyzed from the last four Y Combinator seasons according to the video?
-The video analyzed 417 AI companies from the last four Y Combinator seasons (Summer 24, Winter 24, Summer 23, Winter 23).
What are the top two industries where AI startups are making significant contributions according to the analysis?
-Healthcare and biotech lead with 45 companies, followed by fintech with 38 companies, indicating these are the top two industries where AI startups are making significant contributions.
What percentage of the analyzed AI companies are operating in the B2B sector?
-81.1% of the analyzed AI companies are operating in the B2B sector, showcasing a strong confidence in B2B from investors.
What is the primary use case of AI across all industries according to the data from Y Combinator?
-Automation is the primary use case of AI across all industries, with 69.1% of companies building assistance to help humans deliver faster, and 31% building entirely AI-driven automations.
Which industries are identified as lagging in AI adoption and presenting opportunities for first movers?
-Manufacturing, agriculture, energy, and retail are identified as industries lagging in AI adoption and presenting opportunities for first movers.
What percentage of the analyzed AI startups are leveraging generative AI technologies?
-78 companies, which is a significant portion of the analyzed AI startups, are using or building something using generative AI.
What advice does the video give regarding the focus areas for someone looking to build with AI?
-The video advises focusing on the B2B sector, underserved industries like manufacturing and retail, leveraging generative AI innovatively, and addressing ethical concerns such as data privacy and AI safety.
What is the typical background of Y Combinator-backed AI startup founders according to the video?
-Most Y Combinator-backed AI startup founders have strong technical expertise in fields like computer science, software engineering, machine learning, or data science. Many have educational backgrounds from prestigious institutes or prior work experience at leading tech companies, and a significant portion have co-founding teams.
Outlines
🤖 AI Project Selection Criteria
The video script begins with a discussion on the criteria for choosing an AI project, referencing Paul Graham's advice that the work should align with one's natural aptitude, deep interest, and offer the potential for significant impact. The speaker shares their background in tech and AI, emphasizing the importance of identifying good ideas in the noisy AI landscape. The speaker decides to analyze Y Combinator's portfolio to identify emerging trends and technologies in AI startups, focusing on the hottest industries and sectors, and the skills and backgrounds of successful founders.
📊 AI Industry Analysis from Y Combinator's Portfolio
This paragraph delves into the analysis of AI startups backed by Y Combinator, highlighting the most popular industries and sectors adopting AI, such as healthcare and biotech, followed by fintech and developer tools. The speaker discusses the predominance of B2B over B2C in the AI sector, the focus on application layer over infrastructure due to the latter's complexity, and the common use case of AI for automation and assistance. The analysis also touches on industries lagging in AI adoption, presenting opportunities for innovation, and the variety of AI technologies leveraged by startups, with generative AI leading the pack.
🛠️ Emerging Tech and AI Startup Trends
The speaker explores emerging technologies such as Edge AI, model efficiency, and multimodal AI, noting the low percentage of companies working in these areas according to Y Combinator's data. The paragraph also addresses critical issues in AI, such as data privacy, AI safety, and explainability, and the startups addressing these concerns. The speaker identifies the potential in sectors like blockchain and Quantum Computing, where a handful of companies are pioneering AI integration. The paragraph concludes with insights into the backgrounds of Y Combinator-backed founders, emphasizing the importance of technical expertise and the value of diverse team compositions.
🚀 Recommendations for AI Entrepreneurs
In the final paragraph, the speaker synthesizes the analysis to offer recommendations for aspiring AI entrepreneurs. The advice includes focusing on the B2B sector, targeting underserved industries like manufacturing and retail, prioritizing technical expertise, leveraging generative AI innovatively, and addressing ethical concerns proactively. The speaker encourages viewers to conduct their analysis but provides a concise guide based on the insights gathered from Y Combinator's AI startup ecosystem, emphasizing the importance of standing out in a competitive market and preparing for upcoming regulations and challenges.
Mindmap
Keywords
💡Natural Aptitude
💡Deep Interest
💡Scope to Do Great Work
💡AI Disruption
💡YC (Y Combinator)
💡B2B vs. B2C
💡Generative AI
💡Ethical AI
💡Open Source vs. Proprietary
💡Technical Expertise
Highlights
Paul Graham suggests that work should have three qualities: natural aptitude, deep interest, and the potential for great work.
The speaker has a background in tech and AI, and is seeking the third quality of offering great success in AI.
Why Combinator's selection process is a valuable indicator of emerging trends and technologies in AI startups.
The analysis focuses on the hottest industries and sectors for AI startups, including untabbed potential and ripeness for AI disruption.
Healthcare and biotech lead with the highest number of AI startups, followed by fintech and developer tools.
B2B sector has a significant majority of startups compared to the B2C sector, indicating investor confidence in B2B.
Automation is the most common use case for AI across industries, with assistance and AI-driven solutions.
Industries like manufacturing, agriculture, and energy are lagging in AI adoption, presenting opportunities for first movers.
Generative AI is the most utilized technology among startups, followed by machine learning and NLP.
Most AI startups are proprietary, with a small percentage focusing on open source.
Real-time AI and multimodal solutions are emerging trends in the AI startup landscape.
Startups are addressing critical issues such as data privacy, AI safety, and explainability in response to regulations.
Enterprise solutions dominate the AI startup scene, with a smaller focus on small businesses.
Emerging technologies like blockchain and Quantum Computing are being explored by a handful of AI startups.
The typical YC-backed AI founder has a strong technical background, often in computer science or related fields.
Recommendations for those interested in AI startups include focusing on B2B, underserved industries, and leveraging generative AI innovatively.
Addressing ethical concerns and regulations in AI is a challenging but potentially rewarding area for startups.
Transcripts
what should I build with AI if this is
the question that you are struggling
with watch this video till the end Paul
Graham who is one of the founders of why
combinator the startup school he says
that the work that you do should have
three qualities first it should be
something that you have a natural
aptitude for second it must be something
that you have deep interest in so that
you can you know overcome challenges
that will come your way while you are
working on that project and third thing
is that particular project that thing
should offer scope to do great work now
I have been working in Tech I like
solving problems in the world of
engineering and AI I have deep and Trust
in AI but now it comes down to the third
question the third quality okay what
should I work on that will offer you
know great success and AI as you all
know is filled with noise at this point
partly because of you know there's so
much hype and everybody wants to write
the wave so the whole field is filled
with too many bad ideas at this point it
comes down to how to find good ideas and
in order to answer this question I
thought why not follow someone who has a
track record of identifying and
nurturing successful ideas in the tech
industry why combinator why combinator's
selection process has consistently
surfaced companies that go on to rehap
shape Industries entire sectors and that
makes their portfolio a valuable
indicator of emerging Trends and
Technologies so all I had to do was to
look at the kind of AI companies and
Founders VC is backing specifically I
wanted to learn what are the hottest
Industries and sectors for AI startups
which Industries have untabbed potential
and you know industries that are ripe
for AI disruption U what all companies
and what all startups are solving
uh problems in emerging Technologies
like blockchain or Quantum Computing and
there must be many companies working on
AI safety because we have so many
regulations coming in so companies
working on data privacy AI safety
accessibility explainability
observability those are the kind of uh
insights that we want and lastly you
also want to understand the typical
background and skills these Founders
have so common traits of YC Founders in
order to understand how practically
feasible it is for you to pursue similar
kind of projects now for those who do
not know why combinator is a leading
startup accelerator that provides seed
funding mentorship and resources to help
early stage startups succeed basically
they invest $500,000 Us in each startup
that is accepted into their three-month
program uh in exchange for a small
Equity stake and uh this program aims to
help help these startups dramatically
improve their product help them with
user growth and also uh increases their
options to raise additional funding now
coming to the data collection
process so I collected the data from
YC's startup directory you have more
than 5,000 companies over here that they
have bagged so far I was only interested
in AI companies that to from last four
CS summer 24 Winter 24 summer 23 winter
23 and the tags that I've selected
artificial intelligence AI generative AI
so all of these AI companies are listed
over here and if you look at you know a
sample page from YC U the name is
provided the description over here the
founder details so I've captured all
these along with all the tags that they
have and cleaned the data captured it in
this a table sheet okay so I have
company name description category these
are 4 17 companies that I collected
while looking at a subset of these
companies I have found many exceptional
use cases and in fact uh part of the
data collection process has been uh done
with the help of company called gum Loop
which is backed by YC which was
previously called agentive okay and I
find myself using gum Loop more than I
had imagine now coming to the analysis
I've tried to capture my entire analysis
in this newsletter a article of mine uh
it's called high signal AI uh the link
will be provided in the description
below the first part of the analysis was
to look at the hottest Industries and
sectors that have adopted AI quickly
where people have found really good use
cases and looking at this plot you'll
see that Healthcare and biotech is the
leader here with 45 companies solving
problems in this industry which accounts
for 10.8% of my data of all the
companies that I have collected followed
by fintech with 38 companies 37
companies are building some sort of
developer tools 34 companies are trying
to automate some sort of sales or
marketing workflow and then 18 companies
in education so if you look at these
sample companies I have this study which
is trying to innovate in this education
industry an AI math tutor for every
student so you have these examples in
each and every industry and Healthcare
and biotech is the leader so far
now next you would want to understand
should you build in B2B or b2c the
numbers here are going to amaze you you
have
338 companies out of
4177 building in the B2B
sector 81.1% of the companies are
solving B2B problems only 18.9%
companies are operating in this b2c
sector and you can find a few examples
like gig ml which is helping inter
Enterprises build and deploy large
language models and then b2c you have
Rex pocket pod shortbread these are the
kind of companies and kind of problems
that people are solving in the b2c
sector so these numbers showcase you
know strong confidence in B2B sector uh
from investors uh and uh b2c has a lot
of untabbed potential uh as you can see
that only 20% of the companies are
operating in this particular sector
infrastructure versus application so
this is obvious U as we in traditional
software engineering majority of the
people are going to build in the
application layer that means they are
going to build some sort of application
using the you know underlying
architecture underlying infrastructure
and for obvious reasons infrastructure
layer is hard to build the skill set
required is also rare and investment
requirements are also high so this data
as well coming from YC may not be
representative of the number of
companies operating in the
infrastructure layer uh which obviously
is going to be you know low uh in number
so 14.9% companies in infrastructure
layer and 355 companies operating in the
application layer uh so 85.1% and
14.9% over here automation has been the
biggest use case of AI across all
Industries and there are two types of
automations one is completely AI driven
and another one is uh co-pilots or
assistants so AI assisted human work
that means you are trying to help humans
deliver faster you are automating some
part of the workflow now here 69.1% of
the companies are trying to build some
sort of assistance to help uh humans
deliver faster and 31% of the companies
are building entire AI driven
automations there are companies like
ofon that are trying to automate order
taking at fast food drive-throughs and
there are companies like constructible
that are building co-pilots for
construction teams helping streamlined
projects ideation reduced losses during
you know due to bad data we've seen
which Industries have adopted AI quickly
but what about the industries that are
still lagging which need more and more
Innovation which need more people to
incorporate AI to solve their problems
and these are manufacturing agriculture
energy retail 16 companies only combined
together in these industries so you need
more and more people and these
industries present opportunities for
first movers in AI adoption now let's
talk about the Technologies specifically
AI technologies that these startups are
leveraging now when I talk about AI
Technologies I talk about machine
learning generative AI natural language
processing computer vision uh media
Generation video processing so on and so
forth and take these numbers with a
pinch of solt because a lot of these
Technologies are overlapping so there
may be companies that are using multiple
AI Technologies to solve their problems
and at the top for obvious reasons we
have generative AI because llms are are
advancing pretty quickly 78 companies
using or building something using gen AI
then we have machine learning 56
companies NLP 47 and computer version 18
as I said there will be many companies
which are using both NLP or generative
AI using both machine learning computer
vision so on and so forth moving on to
open source versus proprietary uh now
this data may not be correct because VC
obviously would not want to you know
fund a lot of Open Source companies so
95.7 % 399 companies are proprietary and
4.3% of the companies are building in
open source please note that there are a
lot of companies out there which are
coming out of Open Source projects if we
talk about other Technologies and other
different types of technical trends that
we see in these AI startups Edge AI so
models running on your phone uh Apple is
doing a lot of work in this particular
category but uh when we look at YC data
only two companies .5% of the companies
mentioned that they're solving something
in Ed aai only five companies mentioned
that they're working or trying to solve
something with model efficiency that is
reducing the computational resources
that it needs to train one large
language model so 1.2% of the compan is
focusing on AI model uh efficiency 46
companies are building something with
realtime AI uh that is voice agents uh
primarily multimodal so approximately 22
companies 5.3% appear to be worth
working on multimodel AI Solutions now
as AI is evolving there are a lot of
regulations there are a lot of concerns
around data privacy AI safety
explainability so on and so forth so
there are layers of critical issues that
are required to be solved now there are
startups that are addressing data
privacy and security concerns 18
companies 4.3% explicitly mentioned that
they're solving something in data
privacy and security sector so cyber
security and data privacy one such
company is Coria there are startups only
five uh so far which are working on
ethical AI or AI safety uh so there's a
lot of potential over there startups
making AI accessible for non-technical
users so there are companies like Creo
which is trying to build internal tools
with AI without coding so no code tools
there are three companies that are
working on explainable a adding more
transparency then we have uh 11
companies solving challenges in climate
Tech uh three companies uh trying to
address issues with bias and fairness
using AI we have ai for small businesses
versus Enterprise Solutions so if you
look at this 70.7% of the companies are
Enterprise solution only 88.9% are built
for small businesses looking at these
numbers we definitely need more and more
people to work on these critical issues
around safety regulation data privacy uh
security and uh there is a lot of
potential to grow uh Within These uh
sectors coming to some hard uh emerging
Technologies which is blockchain and
Quantum Computing now obviously these
Technologies and the fields are so
complex that you would not have a lot of
companies solving these problems but we
have a bunch of them Quantum Computing
there are two companies uh which are
trying to incorporate AI with Quantum
Computing solve something over there
then we have three companies working on
blockchain so conductor Quantum
harnessing Quantum computing to
potentially solve complex problems
beyond the reach of classical Ai and
then we have companies like salio or
kadalo merging blockchain with AI for
enhanced data integrity and
decentralized intelligence so there is
still a lot of potential uh Within These
two sectors specifically I'm I'm more
interested in blockchain how you
integrate blockchain with AI to build
something uh amazing now let's come to
the background of the people who are
willing to work on these problems work
on these startups so typical background
of a YC banked founder here you can see
most more than 75% of the founders have
strong technical expertise in computer
science or software engineering machine
learning or data science especially if
you are an AI founder you have to be
very strong in programming so
educational background most around 20%
of the found ERS are coming from
prestigious institutes like Howard MIT
Stanford Berkeley 25% of them have
previously worked at strong leading tech
companies like Google Facebook meta okay
15% of the founders have prior startup
experience so they have worked on you
know some of their own startups before
and uh 8% of the founders come from
academic research backgrounds phds
postdoc University professors and
45% have co-founding teams so that means
a technical founder plus a business or
operations founder so they together make
uh a good team
24% have backgrounds that position them
to disrupt traditional industry so they
have already uh done deep work within
those Industries and they would bring in
perspective that will help the technical
founder or the business founder to take
it further within that particular
industry now not to mention that if you
have done exceptional work in the past
uh without having these titles or
prestigious institutes in your
background you can still make a mark uh
all you need is you know drive and
showcase some evidence that you can
actually you know uh overcome those
challenges that are going to come when
you will work on something hard
something that y would love to uh you
know invest in now now putting this
entire analysis together has what I
would recommend to anyone but obviously
you know feel free to go through the
entire analysis and build the answer for
yourself I would suggest that you focus
on the B2B sector at this point okay uh
look at underserved uh Industries
manufacturing retail if you know
somebody that's good uh prioritize
technical expertise because you would
need uh technical expertise to thrive in
this business if you do not have that
expertise partner up with somebody who
does then uh leverage generative AI in
an Innovative manner to stand out
because there are many companies which
are just rappers and they'll soon be uh
you know out of business because as soon
as open a releases their next Model A
lot of businesses are going to shut uh
because of that then address ethical
concerns this is a category which is
hard to solve but will have a lot of
potential because there going to be many
regulations coming in new bill are going
to pass and privacy cyers these are
issues that a lot of companies are going
to have to deal with so if you can build
a solution around it yeah you are going
to stand out in that particular category
then so yep that's been it and I hope
that you found this video useful
insightful and if you did please give it
a thumbs up that's the best way you can
help me subscribe and I'll catch you
guys in the next one until then keep
learning and keep building with a
تصفح المزيد من مقاطع الفيديو ذات الصلة
5.0 / 5 (0 votes)