I was really excited about Cloudflare Pages when it first launched, but back then it seemed a bit barebones compared to what Vercel was offering (in terms of the admin dashboard GUI, ease of managing previews, seeing logs and function runs, next/image optimizations integrations, ISR support, etc.). Like all of those things were possible on the Cloudflare system, but would've required manual setup across a few of your lower-level services, as opposed to a 1-click deploy on Vercel. (Which is understandable, since Next is their framework after all).
Is it different (more integrated) these days? I'm a frequent customer of both Cloudflare and Vercel, but would love to see more fully-managed frontend services like Vercel.
To clarify, the Vercel OSS program isn't going anywhere. Glad to hear you've had a good experience with Vercel. If an open source sponsorship is something you want to explore, please shoot me an email - lee at vercel.com.
Our bot protection can help with that :) How can we make this easier? Any other product/feature requests in this space I can float to our product team?
If that's already possible I think there's probably a huge marketing opportunity to break it out into a product and shout about it. I'd imagine there's a lot more people out there interested in that than this.
I worked at Twilio for nearly 10 years and it's hard to overstate what a gift it was to work there and see Jeff operate as CEO up-close.
He created an environment where (at our best) we could have fun doing work that had a real impact, and we could it with people we enjoyed doing the work with. He pushed us to be creative to authentically empower and inspire developers. Wanna build a video game that teaches developers how to code and use Twilio? Let's try it! Wanna build an AI application with Tony Hawk and have Tony Hawk debug the code live on stage? Sure!
And Jeff would always be spending time with developer tools and Twilio's products himself, to the point that he could live code at the drop of a hat to show off what we'd been working on. This meant his own understanding of developers and their problems never ceased to amaze me.
But more than all of that, he was a rare CEO that led with empathy, humility and care.
Thank you, jeffiel. We can't wait to see what you build next.
Because Jeff wanted to be a celebrity CEO, and it was to the detriment of the company.
All of Twilio's growth (as a public company) happened on George's watch.
I don't think it's a coincidence that well after George left, on Jeff's watch - they had to do 3 different layoff for a total of ~24% of their entire public company being let go.
Note: former senior level Twilio employee, posted anonymous for obvious reasons.
For a public company, the CEO has to be a public CEO. Everyone takes on a different persona (some sell themselves as business geniuses, others as creative geniuses, and others as someone you wanna grab a beer with). Some really like it, and some do it begrudgingly. But at the end of the day, you can't be the same CEO of a public company as you were when you were private.
I've never known Jeff to hobnob with celebrities, travel in luxury or abandon his company. I'm not sure he's a "celebrity CEO" as much as he just became a public one, although I don't know him personally.
(FWIW, this is the reason I'd never want to go public. I personally don't like the system, but that doesn't mean it's not how the system works.)
Only if you pretend that the CEOs job is purely for raising money.
Nominally, CEOs are supposed to run the company and help brew the culture that makes that company productive.
We've grown a bunch of celebrity CEOs because people stopped caring about how companies worked over the last 10 years of low interest rates. But now that it's harder to raise money, it's time to see these guys run the companies they built.
Everything you said here is true, but I don't think it applies to Twilio or Jeff.
I don't think you can accuse Jeff of not caring about company culture. He built one of the defining dev cultures out there. Yes, this have changed as they've grown, but I've always found the people and offices of Twilio to be warm, productive and smart.
And Twilio was forged in the hardest time to raise money ever... 2008. If anyone knows how hard it is to build in a climate without easy money, it would be Jeff.
Memories are short, but Twilio was built when doing what seemed obvious - add an API to telecom services - was really hard. It's what Stripe is (doing, not done) to banking & payment, what countless companies have failed to do with Healthcare... I met Jeff in the early days and the fact that he (a) managed to stay the CEO through this growth & timeline, and (b) gave few enough sh!ts to maintain a hacker dev mentality makes him a notable standout. I'm not surprised that lots of people with legitimate viewpoints and perspectives DO/DID NOT like him or his approach. To me, that's a feature; we need less CEOs and senior executives who try to be all things to all people.
I was tasked with taking over some of jeffiel’s code bases around the time period we are discussing here. He perfectly skirted the line of telling me what he was thinking while writing stuff and simultaneously staying out of it and letting the people he hired take the reins. He spent most of his time interacting with Cxx and Director level people to set the tone and direction of the company, while still taking the time to know what ICs were doing so that he could usefully answer questions and use our products.
IMO, you and the GP are onto something with the term "celebrity CEO". In Jeff's case, he seems like a celebrity CEO for his employees/developers, who we identify with, though not for his customers, who we are largely not.
His customers probably expected better products, whereas his developers expected another round of amazing perks and fun work.
It's also a double edged sword where greater visibility leads to greater scrutiny. Galen Weston Jr stepped into the public eye much more during the pandemic as the face of Loblaws (Canadian grocery chain), and although it originally was welcome, he ended up outing himself as just another entitled, out of touch zillionaire, and at least in my social circles becoming kind of as punchline.
Why's that? He has been doing the commercials for around a couple of decades, staring long before he was the one doing the casting. Even after taking that role in 2016, the previous commenter points out that the commercials continued to be well received until well into the 2020s. It was a successful formula for a long time.
Indeed, positive audience sentiment has an expiration date. He rode the wave too long, perhaps, but it is hard to know when to stop. All ongoing media productions (e.g. TV shows) suffer from the same problem and question of when to keep going and when to move on.
I didn't actually know that he'd been doing it for a while, but I've not ever been a regular TV viewer. The change from my perspective was the commercials showing up in my Facebook feed and my family group chat joking around about the latest dispatches from old Gal Pal and how we could pick up a shrimp ring for 20% off.
So even if it was a longstanding thing, pushing the content toward a more millenial/gen-z audience is still a new direction and risk.
In the end, the biggest factor was the dissonance between his warm, cozy "all in this together" messaging combined with grocery price hikes and news like this coming on the other side of it all: https://www.blogto.com/city/2022/11/loblaw-report-profit-foo...
Funny thing is that restaurant patronage has declined considerably in Canada in the past year or so. Even with many restaurants now operating at a loss (i.e. paying you to eat) to try and win back customers, they are failing to capture hearts and minds.
If people aren't eating at restaurants, it is almost certain they are getting their food at the grocery store instead. For all the mockery of Galen, it seems the marketing strategy has been unbelievably successful.
That's probably why he continued to cast himself even as the audience focus shifted to millennials and gen-z. Letting someone laughing at you for a few minutes is no big deal. Watching them love you when they show up at your store over and over makes it all worthwhile, I'm sure.
I work in tech and had never heard of Jeff or George. I could name every CEO of Twitter or Google, or the founding CEOs of companies like Stripe, AirBnB, Uber, Coinbase, etc. Companies like Twilio and Unity don't really register.
I've worked on some of the foss tech (infra and ptsn) that twilio has used for almost a decade. Know all about what the company does. Have applied there (to which they never replied). Known plenty that work there.
Ironically my previous partner works there and she didn't know anything about the company prior to taking the job so me not even getting a reply to my application was a jab. I've always seen twilio in a positive light they do some really interesting work.
I live in the city of their HQ. I have no clue who the people mentioned in this thread are either but it makes sense that they'd see the CEOs in a positive light with how much positive I hear(d) about the place.
There actually haven't been that many, and double-checking against Wikipedia I forgot Parag Agrawal but Wikipedia forgot Linda Yaccarino. I divide them up into eras and power struggles:
1. Jack Dorsey. Founding era, founding engineer, forced out when Twitter was ~30 employees because he had no management experience.
2. Evan Williams. Growth era. Another founder, but one who had past experience dealing with hypergrowth, getting companies acquired, and working inside a big company.
3. Dick Costolo. Maturation era. Adult supervision, he's the professional CEO hired to manage Twitter because both of its previous CEOs were really startup people.
4. Jack Dorsey (2). Came back after gaining some political savvy and learning how to launch a boardroom coup. Founded Square in the meantime and got a bunch of experience being CEO of another company. Also famously the guy who shut down the Jan 6th insurrection by blocking the sitting U.S. president from Twitter.
5. Parag Agrawal. Caretaker; Twitter was in trouble by then and Jack Dorsey was either forced out or decided it was better to die a hero than be remembered as a villain.
6. Elon Musk. Bought Twitter in a fit of insanity.
7. Linda Yaccarino. Glass cliff. Installed as CEO so that X's eventual demise is not Elon Musk's problem.
I don't know the right answer here, but there are thousands of public companies. If every CEO has to be a celebrity, meaning that everybody knows them and which company they are the CEO of, it's going to end up like for actors: tier one world superstars, tier two stars, then people you don't hear much about, then people you read their names sometimes in the closing credits, otherwise they are only random faces on screen.
If that's the case, the top layers of the system have space for so many CEOs. Some won't even care to be there, they'll just do their job and take the salary plus bonuses.
Sure. I'd say 99% of people in tech couldn't name the Twilio CEO. I don't think he's really a celebrity CEO, at least not in the way that's being implied about him (as in, I think "public" and "celebrity" are being conflated here).
Jobs certainly was a celebrity CEO, and despite all the criticism of him, I've never once heard anyone imply he was bad for Apple (or its shareholders). Musk is a celebrity CEO, much to the detriment (and potentially benefit? who knows) of his companies.
The CEO’s job is to create value. The answer to your question is “it depends”.
If buyers of their services are impressed with his ability to live code and that moves the needle on results, then yes. If the same can be accomplished by having a non-CEO developer evangelist, then probably no.
There's celebrity CEOs and there's celebrity CEOs.
Without googling, can you tell me the name of the CEO of Target? Sony? Texas Instruments? BMW? FedEx? Mitsubishi? Exxon? Home Depot? Siemens? Procter & Gamble? 3M? Nestle? Broadcom?
I certainly can't.
But I bet you can tell me who the CEOs of Tesla and Amazon are.
Just because whoever-is-in-charge-of-siemens hasn't run a vanity PR campaign to raise their personal profile, doesn't mean it would help the company if they did.
The sum of the market cap of the first group of companies is ~2.068T USD, while the sum of the second group is ~2.270T.
And of course the CEO of Tesla is(?) allegedly the richest man in the world, which is one of the reasons why people know him. And I honestly don't remember the name of the CEO of Amazon since Jeff Bezos stepped down.
lmao Teslas market cap is such a joke. 750bn, 10x higher than BMW while I see tens of millions more BMWs on the road and the BMW brand has astronomically more staying power as brand recognition. Wild. Nobody would be surprised if Tesla flopped but we know BMW, a company over 100 years old won't flop.
I can't imagine being offered a choice between owning (the company) Tesla or BMW and picking Tesla.
edit: Tesla did 4.2% of total market in 2023, outpacing VW, Subaru and BMW. They did 25% more sales in 2023 than 2022. They sell 500-700k/yr right now with 2023 being 660k. BMW delivered 550k cars worldwide.
Shhh you must not rain on techbro's fanatical obsession with fantasy market caps. It's the only aspiration validating their particular niche of career.
This is not specifically about Jeff in any way, but no, it's not a CEO's job to be a celebrity. It's their job to allocate capital in a superior way (relative to the investors desires ie how much beta, and to marginal opportunities elsewhere in the market). If they cannot produce a better Risk Adjusted Return then they simply should be returning the capital, not being a rockstar and growing.
I've seen celebrity CEOs in non-public companies a lot more than in public companies. There is a difference between being the public face of the company and being a celebrity CEO.
George Hu cut his teeth under Benioff during the most meteoric rise of both Salesforce and Benioff's celebrity. I would be very surprised if Jeff's desire to be a public facing CEO would even cause George to notice, let alone bristle at.
It's more likely that someone as smart as George might have just looked at the fundamentals of the business and decided it could not keep this up over the long term.
or... it might have just been epic timing too. Probably.
I don't have any direct knowledge, but the gossip flying around internally at the time was that a big part of George leaving Twilio was because he wanted the CEO role, but knew Jeff wasn't going to leave and give it to him.
That was gossip, of course; sure, George may have decided back then that -- even with all the growth he drove -- Twilio wasn't going anywhere long-term.
The fact that George has not taken on a CEO role since tells me that it probably wasn't a driving factor for him. That's conjecture though. I think that role would be available to him at a lot of strong companies if that was his goal.
Being a celebrity CEO makes sense when interest rates are low and your main job is to keep convincing investors to pump in money. Of course when interest rates are high, and your main job is business and company fundamentals that's a lot less useful. I'm seeing a lot of startups (and public companies) with CEOs like that floundering right now.
I don't know the specifics of these two CEOs, but those were two very different eras financially in the tech sector. To nail all of it on just a CEO's door feels reductive, at best.
Please. George joined the company when it was damn near a decade old, and already a publicly-traded company. It's embarrassing to ascribe anywhere near the agency to someone like that than to a founder.
People who weren't there don't understand, but the world before "Ask Your Developer" was completely different than the world we live in today where developers are at the heart of how decisions are made, products are built and how the world itself is changing.
I joined Twilio in 2012 and I saw first hand how pivotal Jeff was in supporting the idea that in order for a platform to be successful it needed to treat developers like first class citizens, and not simply the recipient of a "directive to integrate" from the CTO.
This idea wasn't simply a fist bump to devs for cool points, it was a recognition that developers themselves held the key to innovation and creating radically new things. This culture led to entirely new ways of building API docs, designing developer dashboards, creating developer events, and so on.
So many of these things have become mainstream and the new baseline for platform companies that people forget how unlikely it was in the beginning and how hard Jeff had to fight to keep the company from treating developers as a means, and not as an end. And I think all developers who are happily hacking away on a free tier of a cool API with excellent docs and a vibrant developer community should tip their hat to a person who helped make this normal and influenced a generation of leaders to do the same.
> Wanna build a video game that teaches developers how to code and use Twilio? Let's try it! Wanna build an AI application with Tony Hawk and have Tony Hawk debug the code live on stage? Sure!
These are the strange effects of ZIRP and infinite QE. Many companies never had to care at all about profit and could just do things "for fun", and still see valuations skyrocket as long as they hired more people. What a time.
I think this is a bit reductive. Sure ZIRP created an environment where people could get away with trying crazy things, but there was still an expectation of growth in pursuit of some longer time-horizon market dominance. These kind of things may be more risky, but they still can be justified in the pursuit of market awareness and growth. Also, when you're building software products, there's real value to engaging the creativity of the team.
The fact that the music continued so long and led to incredible valuation inflation over the last 15 years made the excesses look a lot worse, but even in a more sane financial environment I'd argue you'd still want to do these things to be competitive in an attention-economy, just maybe not to the extremes we saw.
In my 32 years in the industry the best performing companies always did these things regardless of the macroeconomic climate. It’s when you ceded leadership to the accountants and shifted gears to maintenance mode that companies stopped growing. You never did these things for fun, you did these things because they had some knock on effect that made it potentially worth it, and in a portfolio of many small bets some would blow up into your next big product. Even the ones that seem far left field they created an environment where people felt allowed to dream at all, and their knock on effect was in creation of a risk taking culture.
Blaming low interest rates on taking risk and doing highly speculative things, and investing in a culture that values that by funding off the wall stuff puts too much emphasis on the capital markets. If the company was a steel producer, fine. If it’s a company that essentially captures the stuff of dreams and produces a service that executes thought stuff in machines, you have to decouple your R&D from capital management to the extent your free cash flow can accept it.
Note of course low interest rates and other capital market looseness will create more opportunity for this behavior, and not every company doing these things is doing some is a well managed way. But they don’t have to - that’s the magic of capitalism. Through a Darwinian process only those who strike magic win and survive and the others recycle and try again. This feels like a feature not a flaw of zero interest rates - it creates enormous value by virtue of incentivizing taking risk at a time of high innovation. There are times in history where innovation rates were very low, and zero interest rates would just incentivize broken behavior. But at a time when almost all growth is coming from innovation, maybe loose credit is smart. ml
> In my 32 years in the industry the best performing companies always did these things regardless of the macroeconomic climate.
Agreed. Creative, high performing people don't do well in a work environment that's structured like a regional bank.
Valve Software -- incredibly profitable, high performing and private company, did fun things like hire economists (and then later Greek prime minister) to study and simulate virtual, in-game economies [0].
Blizzard Entertainment had giant statues of orcs made and fan conventions before their downfall, ya know, in the ZIRP macro economic environment.
Google allowed their engineers 20% time. Before ZIRP.
I once heard that this was more like 120% time? Meaning, 'extracurricular activities' on company infra was allowed and encouraged, but not at the expense of regular productivity/output.
what about the hardware that significantly leverages their non-game software? I think the full ramifications of what they're doing with mobile gaming have not yet been felt.
> In my 32 years in the industry the best performing companies always did these things regardless of the macroeconomic climate.
I think the question is more about whether doing these things caused the company financial success, or whether the company's financial success caused them to do these things. It strikes me as plausible that it's almost always the latter, even if the company's financial success isn't attributable solely to the macroeconomic climate.
Usually the issue is you can’t encourage creative risk taking with structured austerity. Structured austerity is about improving operational efficiency and there’s a place for that. But at companies that survive and thrive on creative risk taking, giving the reigns to the CPAs kills the culture.
I think it’s usually a sign of success that to protect the golden goose you stop taking creative risks and focus on operational efficiency.
Macroeconomic conditions really matter a lot more for capital intensive enterprises like manufacturing, refining, real estate, etc. Most creative / R&D based companies are much less cost of capital sensitive and frankly low interest rates matter a heck of a lot less for their planning and operations.
> In my 32 years in the industry the best performing companies always did these things regardless of the macroeconomic climate.
My anecdote, FWIW, is that some of the worst performing startups do these tricks too.
There's something about flashy events and boondoggles that sound good on LinkedIn that draws bad founders into spending waaay too much on parties and fun activities.
Stripe obviously isn't in that category, but never assume that because a company spends a lot on parties and events that they're doing well.
> My anecdote, FWIW, is that some of the worst performing startups do these tricks too.
I absolutely agree with this. Monkey see, monkey do.
I think the big winners who make it "fun" are the exceptions that prove the rule. Whatever they really have that leads to success (be it simply luck, timing, connections, market fit, whathaveyou) is a lot less visible and harder to replicate than the lazy, obvious stuff like "make the workplace fun for developers" which any wannabe can emulate.
My wife is not in tech (and so enjoyed Silicon Valley quite a bit less than I did when we watched it together).
Every so often, she'd express frustration about the "over the top writing" in Silicon Valley. Way over half the time, I could tie whatever it was she was complaining about to some concrete story from our industry.
I read a great article about the making of the show and there were many real-life stories they couldn't include because they felt people wouldn't believe it was real.
I think their go-to example was meeting up with a Google X person in their offices and that person getting huffy about something or other and then trying to storm out. But then they had trouble with their keycard so there was that awkward pause because they couldn't just storm out and slam the door. Best part was that they were on inline skates (and rocking a pony tail!) so it was a totally Mike Judge perfect moment - weird silent tension as pony tail skate master is trying to beep himself out of the room.
I've definitely seen and heard about lots of strange things in SV that people simply wouldn't believe if they heard it.
My wife thought the show couldn't be reality. I told her I'm confident that almost everything you see in this show happened in SV to at least someone, just not the same company/people in a short amount of time.
It's like if you took every story in SV history and compressed it to one company in 6 years.
> In my 32 years in the industry the best performing companies always did these things regardless of the macroeconomic climate.
What companies?
Some of the best performing public tech companies of the past 30 years include Amazon, Apple, Netflix, Nvidia, and Microsoft.You can get some variety in this list by varying the exact start and end dates, but certainly the first 4 are going to make most lists. Others like Google and Facebook handily beat the total market.
Certainly a lot of companies that never went public (Instagram, WhatsApp, Credit Karma, etc...) performed very well for their investors when they got acquired. And some companies that since going public haven't been blockbuster successes, like Snowflake, still were probably very lucrative for their VCs and other early investors.
Anyway, I don't feel like most of the really great performing companies that come to my mind are particularly "fun," but maybe I maybe I'm not paying attention and/or don't know because I don't work at them.
Great comment. ZIRP is endless stock buybacks and $boomer_rock_band or $expensive_celebrity speaking at your conference. Those are capital owner and executive benefits and don't drive the product or curate new customers.
Tony Hawk at this event is just a marketing stunt and his celebrity can be beneficial to drive engagement, impress potential customers, keep existing customers happy, help with recruiting, etc. Those stunts can get incredible attention. Look at how common celebrities in advertising and endorsements are.
Stock buybacks, having the Moody Blues play your annual meeting or the Rich-dad-poor-dad guy give a speech to execs and play down to their biases doesn't drive marketing or the product. Its what execs enjoy personally and burn through money for these entitlements. Instead this money should be used to give raises to the working class.
Also I'd substitute accountants for MBA's in your comment.
Share buybacks are a capital return to investors. This is a sensible and desirable thing to do in a lot of macro environments. ZIRP may increase the instances of borrowing money to do share buybacks, just like it would increase the propensity to borrow money to do any other valid corporate activity. There were share buybacks in the 80s and 90s when interest rates were quite a bit higher than today.
Adjusted for inflation, there were more buybacks in 2000 than in 2015 and interest rates were quite a bit higher in 2000 than in 2015.
Were companies in other sectors besides so-called "tech" able to disregard profit and just keep hiring as a result of ZIRP and QE. If yes, what are some examples.
QE and ZIRP are deflationary policies because they reduce the amount of interest payments (which are no different from any other transfer payment) paid by the government.
No, they are inflationary policies; fear of deflation was one of the factors motivating them.
> because they reduce the amount of interest payments (which are no different
That's...mostly irrelevant. Yes, government interest payments would be inflationary because some of them go back into the domestic economy, but flooding the domestic economy with money through loose monetary policy is not net deflationary; the reduction in money pumped into the economy in interest payments is far less than the amount of money that is pumped into the economy to effect that reduction.
QE simply swaps one form of government liability (treasuries) for another (reserves).
They pay less interest on reserves.
The effect on the economy of banks holding reserves rather than securities is otherwise negligible because both securities and reserves can be used to satisfy any liquidity requirements, so reserves and securities are functionally identical.
That’s why it’s so dumb that people get in a flap over the idea of the Treasury issuing reserves “out of thin air” but they are fine with the Tresury issuing new securities “out of thin air”.
No I don’t, there was potentially some inflation due to fiscal stimulus but mostly due to supply chain issues. The continuing inflation is caused by interest rate increases, ie. the interest rate sets the inflation rate.
If you were right, Fed policy would drive the nation into an unbreakable positive feedback loop (irrespective, really, of preexisting conditions because the conventional economic levers would always drive the nation the exact opposite way intended) until conventional understanding of monetary policy was abandoned.
The fact that this hasn't happened pretty clearly disproves your understanding: the conventional understanding may be wrong in some way, but its not inverted.
Interesting. I tried to create a Twilio account 2 weeks ago and it was the worse developer experience I've ever had by far.
Like seriously.
They first make you deposit $20 to get started. THEN, they tell you to talk to customer support to activate your account (customer support took forever to reply).
Oh and then I had to fill out like 3 applications and my text messages are still randomly getting blocked - I have no idea why.
To be clear, this is literally just me using Twilio to text myself to test out the API. I'm not some spammer or something. It was impossible to get started.
This is industry self-regulation. See A2P 10DLC. Dealing with it at scale has sucked. If you want to send SMS, it's the price of admission everywhere--not just at Twilio. I know you said you wanted to send yourself an SMS, but you can test without sending using test credentials and magic numbers: https://www.twilio.com/docs/messaging/tutorials/automate-tes...
it is not just SMS, I just recently set up some AWS SES for sending emails (never done this kind of thing before) and boy is it a nightmare to get emails to not be flag as spam these days.
Like SPF, DMARC, DKIM. What the hell how many layers of security do I need to go through to get some emails delivered from a domain I own?
Strongly seconding this (especially as someone who was backstage with Jeff and Tony right before that demo). What's less visible is the impact and influence Jeff has had in enabling leaders and founders to do the right thing. So much of the pressure and industry influence on CEOs is to be short-term thinkers, to ruin their user experience, or to be harmful to workers.
I've seen, both personally and through others who've relied on his example or mentorship, Jeff be a force for good for those who were trying to do right while leading companies. And to even be someone to call in (or call out) a mistake when you make it — which a lot of bigger-name people are way too cowardly to do. He's also spoken up for his city, and for vulnerable communities, when it would have been so easy for someone of his stature to just stay silent.
It's obvious that Jeff is a brilliant technical innovator, but he is just as much of a pioneer of thoughtful and conscientious leadership in an industry that's become woefully bereft of it.
I'm having trouble parsing this against what Twilio does. Are they talking primarily about the Segment acquisition? I primarily know Twilio as an SMS gateway.
Twilio already uses aggressive sales strategies. They make you deposit $20 to test out their API AND THEN they tell you to get approved by talking to customer support.
I was unable to get approved so I guess I'm out $20 for no reason. They have the worst dev experience I've ever seen.
Probably as good at delivering shareholder value as anyone. Probably just didn't want to play ball when investors started asking him to strip out all the value to the customer and put the screws to the employees and purge anything and everything that might be mistaken for humanity and goodness. They love to do that shit and then pretend their new hollowed out strip mine is "lean".
I had the privilege of working with Miguel at Twilio. His work had a meaningful impact on our goal of empowering developers around the world and I’m unsurprised to see such a thoughtful take from him has he departs. A huge loss for Twilio but I can’t wait to see what he does next.
Ok. I haven't been able to reproduce this issue (I know "it worked on my machine" is the worst answer. But I did do a fresh setup on two other machines to see if I can recreate so it worked on three of my machines?).
If you want to drop me an email (my email is in my profile), I'm happy to help find time to debug live and I'd personally love to find out what the issue is. And/or hook you up with a t-shirt :)
Digging in, this specific error looks like it is happening when LangChain is trying to strip out new lines (you can see the code in LangChain here: https://github.com/hwchase17/langchainjs/blob/main/langchain...). For some reason your documents or embeddings are undefined. But I've been unable to create a scenario where that happens.
I wondered if perhaps your OpenAI API Key wasn't set correctly and it was causing the Embeddings to fail, but I tried both deleting my key and setting to bogus key and the error was different.
Any chance you switched out the URL we were pulling from? That didn't cause the same error for me either, but thinking through what else it could be.
Sorry to not resolve, but please feel free to drop me an email. I'll keep trying to reproduce as well, and will jump back in this comment if I do (and find a fix).
The reason was that I was using json for the input.url input.query - but for some reason, even thought it's just plain strings, it has this error, when i use query strings it just works.
Post author here. I think that's a great description of a "why" for this kind of setup.
I personally feel like with a lot of these emerging AI tools I still find myself still in the exploration phase. Trying to understand how they work and what they do, to then wrap my ahead around why I use them. Some of this post is a look into how I've been wrapping my head around the what and how of LangChain. Doing a follow-up digging deeper into the "why" and what benefits folks may see from this kind of setup is a really good nudge for me. Thanks!
Using LangChain definitely shifted how I was building with language models. It took me a couple days to really wrap my head around the possibilities, so if you feel similarly when you're getting started don't worry.
If you have any specific questions, or want to chat with someone else using it, please don't be shy. My email and twitter handle are both in my profile. I truly love talking about all of this and hearing what other folks are working on.
Hey Jason - Author of the post here. Thanks for the catch, updating the post now. If you drop me an email or twitter DM (details in my profile) and I'd love to send you a t-shirt as thanks for catching this.
You're right, those aren't necessary. That was a setting on our blog CMS that was enabled with a recent upgrade. This is now disabled (it may take a bit for cache to clear and you see these disappears). Thank you for calling this out!
Does anyone know what happens if you do transfer learning in addition to scaling? It feels like people used to use transfer learning in lieu of scaling and I haven't wrapped my head around how they work together.
And we have an OSS sponsorship program: https://blog.cloudflare.com/cloudflare-new-oss-sponsorships-...