Ex-Googlers Discover That Startups Are Hard 61
Dozens of former AI researchers from Google who struck out on their own are learning that startups are tricky. The Information: The latest example is French AI agent developer H, which lost three of its five cofounders (four of whom are ex-Googlers) just months after announcing they had raised $220 million from investors in a "seed" round, as our colleagues reported Friday. The founders had "operational and business disagreements," one of them told us.
The drama at H follows the quasi-acquisitions of Inflection, Adept and Character, which were each less than three years old and founded mostly by ex-Google AI researchers. Reka, another AI developer in this category, was in talks to be acquired by Snowflake earlier this year. Those talks, which could have valued the company at $1 billion, fell apart in May. AI image developer Ideogram, also cofounded by four ex-Googlers, has spoken with at least one later-stage tech startup about potential sale opportunities, though the talks didn't seem to go anywhere, according to someone involved in the discussions.
Cohere, whose CEO co-authored a seminal Google research paper about transformers with Noam Shazeer, the ex-CEO of Character, has also faced growing questions about its relatively meager revenue compared to its rivals. For now, though, it has a lot of money in the bank. Has someone put a curse on startups founded by ex-Google AI researchers?
The drama at H follows the quasi-acquisitions of Inflection, Adept and Character, which were each less than three years old and founded mostly by ex-Google AI researchers. Reka, another AI developer in this category, was in talks to be acquired by Snowflake earlier this year. Those talks, which could have valued the company at $1 billion, fell apart in May. AI image developer Ideogram, also cofounded by four ex-Googlers, has spoken with at least one later-stage tech startup about potential sale opportunities, though the talks didn't seem to go anywhere, according to someone involved in the discussions.
Cohere, whose CEO co-authored a seminal Google research paper about transformers with Noam Shazeer, the ex-CEO of Character, has also faced growing questions about its relatively meager revenue compared to its rivals. For now, though, it has a lot of money in the bank. Has someone put a curse on startups founded by ex-Google AI researchers?
Hahah (Score:4, Informative)
I've worked with Googlers before.
They are fucking annoying and have little of actual value to offer.
Re:Hahah (Score:5, Informative)
Not sure why that was downvoted. I've worked with them before, too, and that's more or less true.
They are fucking annoying
Xooglers are always complaining about the Googley tools they can't use at your company because the tools only exist at Google.
and have little of actual value to offer.
They don't know the commonly-used tools/libraries/techniques outside of Google that everyone else uses. Too long inside Google effectively cripples them as devs anywhere else.
Re: (Score:3)
Re: (Score:1)
Not sure why that was downvoted. I've worked with them before, too, and that's more or less true.
Not me who down-moderated this, but I'll guess that the generic insulting tone of the post overwhelmed the small amount of content.
Re: Hahah (Score:2)
Re: (Score:3)
So hire the ones who build the tools.
"Tool building" is one of the things that kills startups.
Successful startups use existing tools to build products.
Relevant XKCD [xkcd.com]
Re: Hahah (Score:2)
Re: (Score:2)
Xooglers are always complaining about the Googley tools they can't use at your company because the tools only exist at Google.
To be fair, the tooling at Google is amazing. It has some odd downsides and quirks, but it's generally fantastic. For stuff that runs in the data centers, anyway. App development (e.g. Chrome) and other OS development (e.g. Android), not so much, though that's mostly because those other things are open source projects and it's necessary that people outside of Google be able to build and hack on them.
They don't know the commonly-used tools/libraries/techniques outside of Google that everyone else uses. Too long inside Google effectively cripples them as devs anywhere else.
And anyone outside of Google is "crippled" when they join Google, for the same reason. It takes time to g
Re: (Score:2)
It's generally expected that anyone joining Google will need a year to become fully productive.
That learning curve is so long that it's unlikely the Google tools are very intuitive.
Re: (Score:2)
It's generally expected that anyone joining Google will need a year to become fully productive.
That learning curve is so long that it's unlikely the Google tools are very intuitive.
Intuitive is often the opposite of efficient or effective. Tools for casual or infrequent use need to be intuitive, because investing time in learning them would be wasted. But for tools you use daily, for complex tasks, the tradeoff goes the other way.
Re: Hahah (Score:2)
Re: (Score:2)
That's a nonsensical excuse. Tools can be discoverable, easy to use, and effective/efficient. Google can do better.
Most of the tooling in this case is in the form of domain-specific languages, and while the learning curve is steep, there's nothing more efficient for complex tasks than a finely-tuned and perfectly-adapted programming language.
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Because YT is primarily a content delivery service with far less R&D involved in its day to day operations?
It's not the same world (Score:5, Insightful)
Let's be very clear about this, that IT startups following all the right buzzwords are asked about profitability or revenue is a new phenomenon and hasn't been seen in that bracket of the industry in decades.
Re:It's not the same world (Score:4, Interesting)
They forgot to read The Art of Bullshit.
Most startups fail; that's entrepreneurial life. I worked for a few startups whose real goal was to get bought out by a bigger company before the bubble popped or investors realized we were a facade.
I remember the day I discovered our product catalog was a doctored copy of our competitor's. I (foolishly) pointed this out to my boss, and he turned his back to the window, paused, and said, "you will not speak further of this."
I dusted off my resume...
Another time I was told to lie that our product was based on MS-Sql-Server when it was really MS-Access. (We were converting, but it wasn't ready yet.)
I dusted off my resume...
A third place said they were switching to stock options and halving our paychecks. One of the key executives had just left.
I dusted off my resume...
Correction; Re:It's not the same world (Score:1)
> he turned his back to the window
Correction: he turned his back to me to face the window
Running a Business is Hard (Score:5, Insightful)
I've become quite good at the things I enjoy doing. I tried to run my own business and it was a failure, because running a business involved so much more stuff that I didn't enjoy doing and this never found the willpower to be better at it.
The business world is shit. We have to live with itz we have to partake in it, but it is just awful, and being on its front lines is not a place for me.
Re:Running a Business is Hard (Score:5, Interesting)
Re: (Score:3)
Do connections and friends matter more than a quality product?
It's not that one matter more than the other but if you have a great product with no connections chances are the greatness of the product won't matter that much since nobody will know about it. The entire premise of Shark Tank is inventors getting to actually network with people who have the money.
In the case of Bezos he got a ~$250,000 loan from his parents to startup Amazon. If he doesn't get that we are likely in a different world today.
Re: (Score:2)
Did Jeff Bezos have a good product? No he was a middleman. A middleman with connections. How did he leverage 2 day shipping in the early 00s with the Pos
Re: (Score:2)
I was not speaking on the quality of the TV Show Shark Tank but it simply illustrates the point that a good idea is rarely enough. If ideas were good enough the concept of the show would be reudundant.
Amazon was already a popular service before 2 day shipping and originally Prime was an agreement with UPS, my first couple years with the service pre-Amazon logistics was pretty much all UPS but that as you mentioned is somewhat unworkable for your existing package carriers, thus why Amazon is like the size o
Re:Running a Business is Hard (Score:5, Informative)
Amazon was lightning in a bottle. It was being in the right place at the right time. Don't get me wrong it also mattered that it was a viable business plan, that is to say contribution margin generated by the activity crossed over fixed costs at some achievable point; but timing is what let it blow up.
I remember how excited we all were about the World Wide Web at the time. My friends and I were doing things like ordering pizza's just because we could. It did not matter that it took a whole heck of a lot longer to boot up a PC, dial-up, start Trumpet/ppp, navigate to a website, than to just dial "488-8888 for the very best pizza you ever ate" (sorry) ..
Yes the same thing happened with Amazon. Everyone I knew was ordering stuff from them 'just to try online shopping' not because they could not get the same stuff elsewhere without the tediousness of 14.4Kbps. Modems got faster, browsers became more capable, PCs got more RAM, images went higher resolutions and e-commerce moved from this sorta novel twist on a catalog shopping experience to really convenient over the course of a few short years. Amazon by virtue of being 'right there at the start' got the mind share.
Re: (Score:2)
Amazon was lightning in a bottle. It was being in the right place at the right time. Don't get me wrong it also mattered that it was a viable business plan
It also matters a lot that it solved a real problem: finding a damn book on a subject quickly. I enjoyed endlessly wandering the shelves of bookshops, picking a few that seemed right for a deeper perusal, but that was when I was a lot less time poor. But the local bookshop was only ever a small subset of what was available at Amazon, and reader reviews helped immensely in sorting the wheat from the chaff.
Re: (Score:3)
How did Jeff Bezos get all those sales in 6 months? This was a different era. Its hard for zoomers to understand a world before smartphones and everyone on facebook. In the mid 1990s the average person probably had never heard of the internet until the Sandra Bullock movie The Net came out.
Re: (Score:2)
But the local bookshop was only ever a small subset of what was available at Amazon ...
To be fair, you probably could have ordered any book through your local bookshop -- like you can from any B&N now, the huge difference back then was that you most likely had to actually go to (or call) the store and have someone there do the searching for you, perhaps using a phone or catalog, etc... Amazon streamlined that process and made it available from your home PC. Sure the buying was easier but the searching and finding was made much easier. That's true even now with non-book items. I often
Re: (Score:2)
To be fair, you probably could have ordered any book through your local bookshop -- like you can from any B&N now, the huge difference back then was that you most likely had to actually go to (or call) the store and have someone there do the searching for you, perhaps using a phone or catalog, etc... .
Books still in print, yes. And I remember those giant catalogs they'd pull out and plop on the counter. But out-of-print or used books were completely by chance, until Amazon created a central market for them.
Re: (Score:2)
Just to try online shopping? Maybe I was an oddball. I bought things from some websites because the promotional deals they pushed to drive customer acquisition were badly thought-out and deeply tilted in the customer's favor. If I kept my order size small enough, some of them were free.
But Amazon was different; it solved problems for me. I needed a book that was 20 years out of print and in only six libraries in the United States. I found a copy on Amazon and had it in a few days. That was valuable, a
Re: (Score:2)
Makes you wonder how Jeff Bezos sold a half a million dollars worth of books online in a 6th month time period just starting out in 1994/1995 when my high school was using gopher and AOL was bigger than the WWW?
This one is a valid question because I don't know a lot about Bezos in the beginning. I do know he comes off like a giant prick now, but that's possibly caused by his tremendous success. Though I do sense a correlation/causation argument, there does tend to be a trend of "success" === "massive asshat."
Do connections and friends matter more than a quality product?
Having been a silent observer through nearly 40 years of one particular industry and watching the machinations happen in real-time? Yes. Absolutely. I mean, you need some quality to the initial offerings, but
Re: (Score:2)
Makes you wonder how Jeff Bezos sold a half a million dollars worth of books online in a 6th month time period just starting out in 1994/1995 when my high school was using gopher and AOL was bigger than the WWW? Do connections and friends matter more than a quality product? Why did Zuckerberg make it and the Winklevosses didn't? Nvidia was given $20m as a startup from a friend of a former boss, then another $5m lifeline from a friend when Nvidias quadrilateral instead of triangle approach bombed and Sega didn't want to use their chip in the Dreamcast (but paid them anyways because of friendship). That kind of nepotism would get a SuperMicro type report from Hindenberg research today.
Do Bezos, Zuckerberg, Huang, etc. have some inherent ability that determined their success? Or was it luck amongst the shifting, unpredictable winds of economic and technological change? Did Bezos envision AWS or simply have underlings stumble onto AWS? Did the Google guys envision selling ads when they came up with a better search algorithm? Or was their brilliance of a latent variety that was hiding in wait for the opportune economic and technological circumstances?
As a society, we confer accolades on
Re: (Score:2)
I believe the trick is similar to what worked for Apple: Woz was the technical whiz and Jobs dealt with all the business tango needed, being an expert in persuasion and BS.
The problem is getting a partner with matching chemistry, as partnership differences in opinion are a common reason for failure.
My dad once tried a startup with 3 partners: his own dad and another dude. The other dude squandered the money on a fancy office and big desk when it was supposed to be a garage-only endeavor. It's a shame becaus
Re:Running a Business is Hard (Score:5, Insightful)
not true (Score:5, Interesting)
Startups aren't hard. What you need is an actual product. Something you can 1) build and 2) market.
But the business model of modern startups is just to go round after round of financing where nowadays they don't even get money. they get a promise that they'll get money.
It's as if investors figured out a way not to get scammed by people selling a product that doesn't exist, by giving them money that doesn't exist.
Nowadays instead of investors losing millions in idiots like this, it's idiots like this losing their own money.
That's what you get for being a scammer.
Re: (Score:3)
Startups aren't hard. What you need is an actual product.
I think the key to VC money is telling them what they want to hear, just like any other marketing. And the VCs I've met are greedy, vain asshats. I'm not sure what product has to do with it.
Re: (Score:1, Interesting)
I don't know if this site will let me post a comment right now, I think I used my 2 daily comments.
Building something that someone will find useful enough to pay for it, to pay for it, while you don't have enough money for support, you are the one or one of a very small group of people building this stuff, this is hard.
You need to find a few clients right after releasing a somewhat useful first version of whatever it is you are offering. If you don't find clients in the very beginning, this may not go well
Re: (Score:2)
what you say is true, however, ftfa:
i think very, VERY few companies need 220 million to build a product.
and if you're struggling because you only raised 220M, i think you didn't really have anything to sell in the first place.
Re: (Score:2)
My friend, every fucking startup I've worked at (and it's like 5 at this point because i work for a software talent agency), has been nothing more than a glorified CRUD.
People into "Radar and communications systems, missiles, avionics maybe" aren't into VC territory. This is defense money which is a completely different beast.
I think you're out of touch with what the whole "startup tech" thing actually means nowadays.
Re: (Score:2)
Maybe it's true that startups aren't hard. Making one that can survive financially, is very hard. About 75% of new businesses ultimately fail. https://www.investopedia.com/f... [investopedia.com].
Re: (Score:3)
A lot of tech startup products are trivial in the technical sense. I'm talking about digital middlemen apps - Uber, AirBNB, Instacart, etc. They just provide a platform where non-employee people get a tiny cut for doing the bulk of the work. The end user pays a high markup but almost all of that goes to the parent company that is still bleeding dry.
The only thing that makes this viable is pouring huge amounts of money into advertising and buyouts and expansion and survive long enough for everyone else to
startups are hard! (Score:3)
Re: (Score:3)
Maybe the AI bubble (Score:5, Insightful)
is at full stretch and we're finally seeing investors looking harder at what they're buying.
Severe Acute Bluediniosis (Score:4, Insightful)
Maybe a better article title would be... (Score:2)
"AI Bubble finally bursting?"
Yes, Startups are hard. But it helps to have an actual useful product idea and business plan.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
> it helps to have an actual useful product idea and business plan.
Not necessarily. I've worked for startups whose real goal was to get bought out by a bigger company before the bubble burst. Once can "surf BS waves" if their timing is good...or lucky.
Re: (Score:1)
Naivete rules the day (Score:1)
Most people don't understand that, once you sell a bunch of shares to a bunch of vulture capitalists, they are in your asshole every single minute of your life dictating what you do with your business, in the best interests of their profits and not your, or shall I say their business.
A vulture capitalist's only goal is to sell your company at a profit, as quickly as possible, and for as much as possible. It's to make your company APPEAR like it is a good purchase, regardless of the underlying fundamentals.
Walking away from 220 Million? (Score:2)
If they can afford to walk away from a 220 Million dollar scam it must still be easy to start an even bigger scam.
Troll-ish headline (Score:2)
Google has almost 200k employees, there's a lot of "ex-Googlers" out there.
The interesting thing was that Google dumped some of the AI talent they accumulated and some of them are currently launching startups.
I don't know about the snarky take that they're "discovering startups are hard". Most startups fail, I don't see any reason to believe that these particular folks were any more optimistic/delusional than the average founder.
Now, their investors on the other hand were a bit questionable. I don't know wh
Re: (Score:1)
The interesting thing was that Google dumped some of the AI talent they accumulated
The first obligation of M&A is to eliminate redundant staff, and that includes 'talent'.
Silver spooners ... (Score:3)
Ex-Googlers Discover That Startups Are Hard ... Has someone put a curse on startups founded by ex-Google AI researchers?
The best description of wildly successful mega startups like Google, Apple, Facebook, Microsoft, ... that I have ever come accross is the title of Robert X. Cringely's book "Accidental Empires". I suppose that when you have spent enough time at Google, et al. in the company of the iconic 'founders' who run those corporations, people who have come to think they are some kind of business übermensch when in reality a lot of their success is a result of them having threaded a golden path simply by virtue of having been in the right place at the right time out of sheer dumb luck, it is easy to get the impression that growing a startup into a Fortune 500 corporation is easy.
Re: (Score:3)
Ex-Googlers Discover That Startups Are Hard ... Has someone put a curse on startups founded by ex-Google AI researchers?
The best description of wildly successful mega startups like Google, Apple, Facebook, Microsoft, ... that I have ever come accross is the title of Robert X. Cringely's book "Accidental Empires". I suppose that when you have spent enough time at Google, et al. in the company of the iconic 'founders' who run those corporations, people who have come to think they are some kind of business übermensch when in reality a lot of their success is a result of them having threaded a golden path simply by virtue of having been in the right place at the right time out of sheer dumb luck, it is easy to get the impression that growing a startup into a Fortune 500 corporation is easy.
While your comment is absolutely true, I think it's worth pointing out that out of those you listed, only one of them was actually built on a groundbreaking technical innovation: Google. Apple's success has all been about excellent execution in the design and construction of commodity hardware (and software, a little bit). Facebook had nothing novel and didn't execute particularly well, but just got lucky. Microsoft's big start was really about connections; Bill Gates knew Tim Paterson (author of 86-DOS)
That's an oversimplification (Score:2)
Startups are hard, but so is running a company, coming up with a profitable idea or product, and working with other people who believe they are as 100% right and correct about everything - just the same as you.
There's probably even more components
correlation (Score:2)
Are you sure the common element is "ex-googlers" and not "AI startups"?