Pivot and Fail Fast: Most Valuable Lessons During Our 12-Year Journey to IPO | SES AI, Qichao Hu

Name: Pivot and Fail Fast: Most Valuable Lessons During Our 12-Year Journey to IPO | SES AI, Qichao Hu
Duration: 14 min 53 s
Description: Meet Qichao Hu, Founder and CEO of SES AI, who shares his transformative journey from MIT researcher to public company leader. With a BS in Physics from MIT and

EO14:53Added Aug 31, 2026

Meet Qichao Hu, Founder and CEO of SES AI, who shares his transformative journey from MIT researcher to public company leader. With a BS in Physics from MIT and a PhD in Applied Physics from Harvard, Dr. Hu made a few pivotal decisions including transition from academia to entrepreneurship and from traditional battery manufacturing to AI-accelerated material discovery, since founding SES AI to revolutionize the future of energy transition.

Watch on YouTube →

Contributed by 刘嘉琪

Transcript

00:00so I think the one thing that I wish I
00:02had done differently is I wish I had
00:04made these pivots faster sometimes you
00:07have to change your DNA faster but then
00:09because you are who you are and you
00:11don't want to change something that's
00:12quite core to you that takes a long time
00:15at least realize that the things that
00:17you're doing are influenced by who you
00:20are and and by the human aspect for
00:22example we used should focus on drones
00:24and then we had people that were quite
00:26close to drones companies that we work
00:29with and then also they spent a long
00:31time developing the drones cells and the
00:34drones design and when we say now we
00:37pivot to EVS we to let go all that all
00:40the relationships we build with those
00:42people all the work that we did all the
00:45momentum that we build we had to let
00:47that go and that's not that easy it goes
00:49back to Pivot when you make a pivot
00:52there is a risk and the biggest risk is
00:55you are letting go of something it's
00:57quite natural for you to be emotionally
00:59attacked attach to the thing that you're
01:01trying to let go but you have to let go
01:03the sake of growth so I think I wish I
01:05had let go of momentum the pass the DNA
01:09faster so that we could actually pivot
01:12faster my name is CH hul I'm the founder
01:15and CEO of SCS aai we develop Next
01:18Generation lithium metal batteries for
01:20electric transportation both on land and
01:23in the air we also use AI to accelerate
01:26the pace of material Discovery and
01:28improve EV safety we've been around for
01:31more than 10 years we've raised more
01:33than half a billion dollars now we're
01:34listed on the lyc and we were the
01:36world's first to develop Automotive a
01:39sample and Automotive B sample lithium
01:41metal sales and we partner with global
01:44card companies including GM Hyundai and
01:52Honda so when I was a child I was
01:54fascinated by the universe and then in
01:56terms of anime Transformers just like
01:58any other kid that grew up in the late '
02:0080s and the '90s everything that we have
02:03today Once Upon a Time was an
02:05imagination a lot of things in
02:07Transformers for example cars flying
02:09cars space travel nowadays we are quite
02:12close to be living in a society where
02:15you have these things everywhere I was
02:17fascinated by physics show a lot of my
02:19time and MIT in undergrad SCI physics
02:22and then at that time when I started my
02:25PhD in 2007 2008 it was a very special
02:28time in the US
02:33it's the largest lithium ion automotive
02:35battery manufacturing plant in the
02:37country and today the company A123
02:40Systems will officially open their new
02:42plant A123 was at the peak I would go to
02:46these seminars MIT and then just hear
02:49Founders top Executives at A123 give
02:53talks about their experience A123 was
02:56the American hero in American Battery
02:59industry but then in 2012 they could not
03:01compete I think A123 was too early too
03:05advance for its time they had this lfp
03:07technology but back in 2009 2010 2012
03:11the popular Chemistry by the Japanese
03:14and the Korean companies was not LP it
03:17was NCM at that time the OEM like GM
03:20Ford they would make A123 and the Korean
03:24and Japanese salemers compete and
03:26obviously A123 could not compete and
03:28they lost and then they went bankrupt so
03:30founder cosos and professional cosos are
03:32really different founder CEOs are very
03:35technology focused and they really care
03:37because as like your baby but when you
03:38are a professional CEO you don't know a
03:41lot of the the Deep challenges and the
03:44Deep issues and then you get funding
03:45from the government and then you quickly
03:48forget who you are then their
03:50Professional Management forgot that
03:52because they got a lot of funding and
03:53they thought they could expand really
03:55fast they hired 5,000 people in 3 months
03:58build this giant Factory in Michigan and
04:01then it was like from the moment they
04:03announced this build out and the hire 3
04:06months later they had 5,000 people they
04:09had just barebone facility and then 3
04:11months later they had to lay everyone
04:13off growth was too inorganic was too
04:16fast and if you make one mistake and
04:18also they lost focus on quality they
04:21were focusing on management focusing on
04:23just public marketing they had an issue
04:26with one of their I think was like a tab
04:28welding those defects cells ended up
04:30getting shipped to GM obviously that was
04:33not good so they lost the contract to LG
04:35so the A123 Professional Management care
04:37a lot about revenue is a listed company
04:40and revenue is important for your market
04:42cap but they lost focus on the cost it
04:45was huge Revenue with huge negative
04:47margin just it was not sustainable and
04:50the whole thing blew up I watched the
04:51whole thing and I interface with
04:54everyone both the founders Professional
04:56Management back then I remember just
04:58saying to myself don't let it happen to
05:03CSI end of 2012 after A123 went bankrupt
05:06because A123 is here in wam at that time
05:09we were just just trying to raise our
05:11series a we had no money so we went to
05:14A123 because we had known those people
05:17quite well and then their R&D facility
05:20in Wan was was available it used to have
05:23at the Peak 200 people at that time you
05:25had like eight people and then all the
05:27dry rooms all the equipment there were
05:29new sale VR but then everyone's gone so
05:32we went there and we just asked one of
05:34those Engineers hey can you use the line
05:381865 line and help us build build a
05:40couple of prototypes for free so they
05:42did that so that was uh end of 2012 like
05:45we would go to A123 and then just
05:48leverage whatever they had and then just
05:50build prototypes all for free looking
05:52back that really saved us a lot of time
05:55and then a lot of money was a a very
05:57good experience no other battery company
06:00in the US went through that we were the
06:02only company in Massachusetts at that
06:04time that leveraged that A123 and I
06:07think both that experience leveraging
06:10the A123 incubator learning from A1
06:12123's mistake A123 had a good technology
06:16but they expanded too fast they got a
06:18lot of funding and they almost forgot
06:20who they were I think one thing is when
06:23the market is is really good we try to
06:26get as much cash as possible but then we
06:29don't don't invest many others for
06:32example in 2022 in 2023 and also I will
06:36say first half of this year when the
06:38market was was actually quite quite hot
06:41a lot of our peers invested a lot if you
06:43go visit our peers in California in
06:46other places I think you'll probably
06:48like their facility more than our
06:50facility we tend to be pretty pretty
06:52cheap but then a lot of those companies
06:54that invested in these things they are
06:57now shutting it down the market is good
07:00we trying to get as much cash as
07:02possible which is what we did with the
07:04spa in 2021 and 2022 but we don't invest
07:09aggressively as many of our here
07:11especially in CeX because I've seen
07:14companies that spend $100 on cack fire
07:19cell situation will have to sell that
07:20for less than one
07:25cent yes so when you run a company
07:28especially want to become listed there
07:30is a lot of work that you have to do but
07:32you also don't have to do you have to do
07:35a lot of work for the sake of of a
07:37shareholders Market investors but those
07:39actually don't really add value to the
07:42PO Mission I mean I think there are lots
07:45of challenges we F raise seven times
07:49sometimes Market is good sometimes
07:51Market is tough and you have to raise
07:52from private investors from Financial
07:54investors from strategic investors it's
07:56challenging technology is challenging
07:58deal STS challenging these are okay
08:01these are challenges that we have to
08:03deal with I think the biggest challenge
08:05that I hate the most is um is a bad
08:08board so when you're the CEO of a
08:10company there's a board right when
08:12you're private the board is consist of
08:14investors when you have like a bad
08:17investor that doesn't know anything but
08:19wants to get involved could be this
08:21investor has like a timeline and they
08:24need to exit or they have say this new
08:27fund and they have pressure from P the
08:30most challenging and then it's the worst
08:32kind of challenge it's not like a worthy
08:34challenge it's not like technology or
08:36business deals they're challenging but
08:38they're interesting and they are worthy
08:40of your time investment I think one you
08:43need to be as an entrepreneur when you
08:47raise money from investors you should be
08:49very selective in the type of investors
08:52some investors maybe it's a big name
08:55investor uh but then if they are too
08:57controlling you want to avoid that a
08:59dollar from a bigname investor is the
09:01same as a dollar from a less known
09:03investor it's it's really the value of
09:06the dollar and you really want to select
09:08a good board and a good investor
09:11investors that have good chemistry with
09:13you when you have these agreements
09:15shareholder agreements and the Border
09:17agreements be really clear with what
09:19kind of power you have and I think um as
09:22entrepreneur now is not a time to be
09:24nice you have to be really clear in
09:27terms of what power the CEO has and then
09:29what power the board has and it's it's
09:32really not the time to be polite if you
09:34don't give me the power don't make me
09:36the CEO make someone else the COO if
09:38you're going to make me the CEO give me
09:39the power I think that's really
09:41important if you get too complicated
09:44there it's hard to focus especially when
09:46situations get bad when you have to make
09:48pivots if there is not clear and
09:51concentrated power it's actually hard to
09:53make things
09:57happen yes so the current EV Market is
10:00challenging we are in the industry where
10:02the validation process takes a long time
10:05it should be because for someone to buy
10:07a car and then drive that car with their
10:10family that car should be safe it takes
10:12a long time to qualify a new battery
10:16chemistry in cars when you change a new
10:18material when you change a new
10:20manufacturing process when you change a
10:22new design it takes a long time to
10:24qualify this is also one of the reasons
10:27why we develop the three AI Solutions so
10:29that we can accelerate that timeline
10:32actually taking 10 years of seeing how
10:35it behaves under different real world
10:37situations let's take one year and then
10:40take all the data from one year play an
10:42AI model and then have that AI model
10:45tell us how it will behave in the next 9
10:48years so I think the the use of AI can
10:51accelerate the time that it takes
10:54customers and the market to get
10:56comfortable with new Innovations the
10:58three AI solutions that we are
11:01developing AI for science AI for
11:03manufacturing and AI for safety really
11:05start to form a new ecosystem that did
11:07not exist before and this new ecosystem
11:10is now forming because you have more
11:12data from all the gigafactories that are
11:14becoming available and also you have all
11:16these new models that are being
11:18developed large language models New
11:20Foundation models and all these new
11:21models that we are developing when these
11:23two converge the data from these
11:25gigafactories and these new models you
11:27can form this really exciting ecosystems
11:30that never existed before and then we
11:32are working with in addition to the car
11:35companies that we worked with before
11:36Hyundai Honda and and GM we're now
11:39working with computer companies for
11:41example like U Nvidia and and a few
11:44others to really establish very capable
11:47computer centers and also to develop
11:50these new softwares new models that are
11:52tailored to the type of computation that
11:54we're doing so that we can really
11:56efficiently and then quickly compute a
11:59large amount of data from Material data
12:02to manufacturing data to vehicle data I
12:04think we are taking this this Challenge
12:07and then turning this into a very
12:09exciting opportunity where we're forming
12:11a new AI ecosystem in the EV battery
12:15industry that never existed before I am
12:18personally very interested in
12:19transportation related applications
12:22Transportation a that's why EV
12:24Transmission in the air so flying cars I
12:27really believe flying cars will be
12:28everywhere in 10 years worldwide it's
12:31something I'm very passionate about and
12:33I truly believe in when I was a child I
12:36really like the Greek mythologist of
12:39Hermes it was one that brought dreams
12:41and then one application I'm
12:44particularly excited about for the use
12:46of lithium metal is high altitude Pudo
12:48satellites so these are these drones
12:51that fly in the stratosphere so 30,000 m
12:55above the ground and then it's basically
12:57a flying cell tower so you can imagine
13:00these high altitude stud satellites will
13:03fly in remote places and then provide
13:06Wi-Fi provide signals to kids that have
13:09no access to Wi-Fi in terms of the
13:12spread of knowledge the spread of
13:13information to uh people in remote
13:16places I think it's a very exciting
13:19application and also on the material
13:22side we are building bigger and bigger
13:24GPU clusters in the Arctic and this
13:28Arctic Computing Center can power more
13:31and more applications from batteries to
13:34uh for flying cars but also I think is
13:37going to expand Beyond transportation
13:39for example there are many challenges in
13:42climate change you can use this Arctic
13:44Computing Center to address a ocean
13:47plastic waste or to come up with new
13:50antibiotics or to come up with new uh
13:53models the Earth climate SCSI will uh
13:57evolve from a battery company to power
14:01company for transportation and then also
14:04brain for to address climate
14:18change I think as a human it's it's
14:21quite natural to assign an emotion to
14:24success and emotion to failure but
14:26actually there are just different labels
14:28of different States right I don't really
14:31think too much about about these
14:33failures obviously when these things
14:35happen there is a cost emotionally
14:38financially and whatever really in the
14:41grand scheme of things it's just a state
14:44temporary state I'm not that bothered by
14:47those I'm a robot