But it makes me wonder how people have time to invent all this shit when they need to move fast. Learning a new stack every 2 years doesn't seem like moving as fast as you could.
You don't have to pick an exact moment, but there are several clear moments of fraud. The biggest moment of fraud was whenever alameda's FTX account was allowed to have a riskier position than a normal account, including holding a negative balance. This is something FTX specifically allowed and publicly lied about (SBF on twitter said they were treated as a regular account, which we now know is false).
This fraud became worse the more negative the balance became since it became more and more misleading to users to continue to present themselves as a financially healthy business.
What they should have done instead is to not allow alameda to have an account with special rules and negative balances, and to actually operate their business as they claimed. They likely would have grown less quickly or even gone out of business, but sometimes the way you avoid fraud is indeed by going bankrupt instead of lying.
They also could have not lied to users, both about the status of alameda's account, and about the status of FTX's liquidity. Each time they made a statement about their exchange that excluded that materially relevant information was also fraud.
I do say all of this with full awareness that a large fraction of YC startups also "fake it until they make it", i.e. lie about what their product can do, lie about their financials, etc. Most YC startups are also committing fraud on a much smaller scale, and just manage to keep the scale small enough that no one goes to prison.
Hold your horses! What startups are lying about what they do? I don’t get the impression this is the case at all. MVP and move fast tactics are not fraud. Saying your SOC2 compliant or something when you are not probably is.
I know first hand of YC founders lying to (corporate) customers on their business size, revenue, etc in order to secure business. Is that fraud? At best its an unethical and dubious business practice.
It is only fraud if a jury of your peers would think it is fraud, which is probably a higher bar than you think.
Most of the fraud you are talking about also wouldn’t have the explicit benefit directly to the founders the way it did with FTX. SBF took money that wasn’t his and spent it on private planes and luxury apartments and hundred million dollar investments _in his name_.
The average startup founder who is trying to close a sale by exaggerating their business a little bit is not going to have that kind of clear gain, because most founders do not use their companies as piggy banks.
> Most of the fraud you are talking about also wouldn’t have the explicit benefit directly to the founders
The "Frank" startup that sold to JPMC because they lied about active users [1] directly contradicts your conclusion. It happens, probably more than we realize, and founders have an incentive to do it so they can reach a liquidity event to cash out.
There's not necessarily an "exact moment" that they can pinpoint, nor is that their job to do so.
FTX was promising customers it would hold onto their funds and keep them safe, much like a bank. Then they did not. That's the short of it, tat is what they uncovered. The date it "began" was not in scope of the case, just the end result, and the end result was that there was fraud.
> What should he have done instead?
Again, the short of it is: He should have delivered what he promised to customers, or not make those promises to begin with.
1. Sam Bankman-Fried SHOULD NOT HAVE STOLEN CUSTOMER MONEY.
2. Sam Bankman-Fried should have been honest.
3. FTX should have kept excellent records. FTX was notorious for its bad record keeping. Also, its poor record keeping meant it was not sure how much money it had and it was not sure which money it owned and which money its customers owned.
I'm sure there was a long slippery slope to more and more brazen fraud, but for what he should have done instead that's a very very simple question to answer. He shouldn't have committed fraud. All the illegal stuff he did with comingling funds and stuff? Just - don't do that.
When he used customer funds deposited into FTX to gamble through Alameda.
Since customer deposits into 'FTX' went straight to Alameda's bank account, it began pretty much the moment Alameda executed a trade. Everything after that was just more fraud to cover up the consequences of the original fraud.
> What should he have done instead?
Not treat customer deposits like a personal piggy-bank.
When Alameda, before FTX existed, gave investors leaflets saying Alameda had guaranteed 20% returns and it could sustain 20% profits on as many billions as they would get. This was a lie and criminally defrauding investors already.
SBF never did the Kimchi trade (i.e. arbitraging the US/Asia Bitcoin price differences) at scale (it's not even clear if he even managed to do it at all: he claims he did, but he's a pathological liar who kept lying in court).
So Alameda was already a scam. Then FTX was a bigger scam, to keep the Alameda scam going.
Everyone bought into "immutable" "no side-effects" "one way data flow" as this ultimate goal. But never thought what the tradeoff is.
And people get FOMO that they need to use a big thing that they couldn't write themselves.
Whereas if you throw out everything and just build only what you need, you get something infinitely more understandable and simpler, probably with some hand-rolled data-binding thing that you actually understand.
Most people don't understand how their frameworks are working under the hood. They are just relying on a nicely documented API...until anything goes wrong.
Trying to debug React for example is insane. And if you build your own custom store it re-renders 6 times when in dev. When I saw this...I knew React's days were numbered.
Yeah that’s the ultimate interview question from Facebook: “explain why this component calls render() 46 times” and the answer is “it doesn’t, it renders 47 times” and they have to bring in one of the core developers to stop all the head scratching.
Yeah, the last paragraph is the clincher. When it's going wrong, on fire, lagging out your browser, and you're adding console logs in effects to chase what is driving changes you've lost.
> the first, MVP-style, attempt will be using Electron.
Good choice.
If you want to stick to HTML...WebView2 is suppose to replace it, but it assumes your backend code is .NET or C++, instead of JS/Node.js. Microsoft Teams is using it.
If you want to use native Windows UI components from JS code, then React Native for Windows is recommended. Facebook Messenger is using it.
All approaches require C# or C++ modules to be used to interact with the Windows Platform. Or there is: https://github.com/tjanczuk/edge.
The recommended approach is WinUI 3 which would involve C#/C++ and XAML.
The biggest issue for Mac users is the placement of the `command` key on the US ANSI keyboards - which is the layout for pretty much all mechanical keyboards. And the missing left `fn` key.
It requires you to curl your thumb awkwardly when resting on the home row or WASD.
It is between the Z and X, but on MacBook keyboard it is directly under the X so your thumb can rest straight.
The only keyboard I have found that has a similar MacBook layout is the Niz Plum Micro84[1] with dome switches.
There was also the discontinued NuPhy F1 that had a fn key but still had a terrible command placement.
I don't know why more Mac users don't complain about this.
Now imagine working on an assembly line on your feet all day versus your office day job. The people on the assembly line would look at you like the most spoilt people in the world.
I personally hated going into the office, but the morale effect is real.
I think the main thing is people cannot be trusted to honestly tell you if they work better from the office. It's a nicer life working from home but whether you get more work done is debatable. And there is always too much to lose by speaking the truth.
> The people on the assembly line would look at you like the most spoilt people in the world.
It's not a matter of being spoiled or not. I'm talking about what is necessary to get the job done and the elimination of redundant rules that serve no purpose.
You can make that point about literally anything. People in X situation have it worse than people in Y situation, therefore Y situation is actually fine. It's a deflection without any actual merit.
If this were a widespread attitude surely the office workers would be paid less than physical laborers (and, frankly, they probably should be). I do a lot of physical work (that's my two primary jobs) and people don't spend their time fuming over people with soft hands—that's kind of a "loser" thing.
At the same time, that office worker has to take time out of their free time to get exercise while the Amazon driver gets paid to exercise on the job. There's a lot of health risk with not getting daily exercise.
The morale effect might be real for you, but for a large group of workers, their morale has improved since WFH. On my team of 24, only one chooses to work in the office...
Individual morale and team morale are separate. I have no doubt that their individual lives improved a lot (e.g., being able to freely do personal chores), at the expense of team morale (feeling like a part of a team at work).
>>>> The people on the assembly line would look at you like the most spoilt people in the world.
That's been part of workplace culture even when most people were in-office. The "office" and "factory" people had noticeably different working conditions. Moreover, the remote workers were largely invisible -- if we dealt with them at all, it was through their boss, or some kind of ticketing system.
> The people on the assembly line would look at you like the most spoilt people in the world.
And farmers working hard manual labor from before sunrise to sundown, in heat and UV exposure would look upon factory workers as spoilt. So what? Are you against progress? You aren't one of those "I had it hard, so these damn kids should too!" boomers are you?
My issue with every piece of commercial software I’ve used is the lag. Slack, Google Meet, Zoom… All have noticeable delay which make conversation unnatural and painful.
I’ve been using FaceTime Audio to communicate with family on the other side of the world and it’s incredible. Zero latency and great audio quality. It’s an extremely underrated piece of tech. Not sure about FaceTime video as I’ve never used it, but I bet it’s equally impressive.
Amazing that no one else was able to solve this problem considering the demand for such software during the pandemic. Not my area of expertise, but I assume it must be quite difficult.
I can't speak for FaceTime, but I've found much better success with all of these platforms if everyone in the meeting is wearing headphones. A lot of the latency is from echo cancellation, and that turns off if there's no echo.
At my last company, I let everyone on my team expense comfortable headphones with a boom mic, and it made a huge difference on Google Meet. I've found Zoom to be similar. I can't speak to Slack's latency issues.
I'm also a big fan of Tuple[0] (the pair coding app) for extremely low latency screen sharing / pair coding and that was a huge advantage too.
I’ve used these tools on all sorts of network configurations and there’s always lag.
One thing I will say is that this was all within Australia. Surely the packets aren’t going overseas, but perhaps whatever servers they use in AU just aren’t optimised for ultra low latency.
Lag is one of the issues I've never had with any of those services. I think they all have their CDNs built up enough so thats not a real problem for the vast majority of people.
I agree totally. It makes having conversation between multiple people very awkward with many pauses, then suddenly people all talking over each other, then again awkward pauses. It feels like online meetings are way more "centralized" in the sense that only a couple of people speak, and most other people remain silent. There are other technical problems to video conferencing, but lag is one of the biggest.
Agreed, people keep trying to push Teams, Zoom, WebEx and others, and everyone complains about how rubbish video meetings are. Google Meet work, and it works well every single time.
Not that I enjoyed using Google Chat much, but it did have one awesome feature: Click and the you where having a video meeting/call with the people you where just chatting to. The absolute simplicity of either planning or just setting up a spontaneous meeting using Googles solution is fantastic. I really don't get why more aren't using it.
I worked at Google for 4.5 years. If you are video calling into a meeting from MTV where other people in MTV are in person, you are probably not doing real work. It's not seamless and it's orders of magnitude less productive. Meetings take longer and rarely produce useful follow-ups if you're remote.
Mind you, I didn't say it was the same as in person. Just many common frictions of video conferencing -back then- were already eliminated. The room knew which video conference to join, noise cancellation was good, screen sharing was trivial...
> But not being in-person for collaboration/morale is miles apart.
I do agree that collaboration is much more productive when remote.
Nearly all of my exciting office "collaborations" end up feeling great then they're happening, but in retrospect, almost always fizzle out after a small window of time. I like onsites for the energy and socialization, but even then I always find most of the work happens back in the hotel room at night, and real planning happens after everyone is back at home. I can't imagine working that way all the time.
Whereas all of my remote collaborations are well documented while they're happening. Typically we have some kind of shared note taking, and writing code collaboratively, etc. Doesn't feel as social sometimes, but tends to be a much larger impact.
Likewise Open Source software has been largely written by remote people since long before the advent of even video conferencing. Git was originally written with the design intention that kernel hackers could work on a plane (per airplane wifi) without requiring a centralized server to communicate with.
> Video conf still sucks. It's literally the same as 10 years ago.
I still marvel at this. I've been working remote for nearly 15 years now and I honestly don't feel that video conferencing has improved noticeably.
My brother is looking at a new position that does 2d/week hybrid. He is thinking of laying it out this way:
* * W R F
M T W * *
So that he has wed to wed where he doesn’t go in.
Also some workers stagger their schedules and share rent on an apartment that is a “crash pad”, so they can live farther from the metro and then just commute for the two day office visit in one trip.
Pretty interesting seeing how people adapt to this.
Audio calls also still suck. Commercials in the 90s advertised "crystal clear audio quality" - they stopped talking about it but it never materialised.
> Audio calls also still suck. Commercials in the 90s advertised "crystal clear audio quality" - they stopped talking about it but it never materialised.
So this is a funny one because its definitely solvable but it's kind of a classic principal agent problem. There's a number of things that are going into 'clear audio quality'.
1. Receiver's audio output quality - This is rarely if ever the problem. People usually use their speakers/headset to listen to stuff besides corporate meetings, so they're motivated to have something at least good enough that they can enjoy their music, talk shows, whatever.
2. Bandwidth allocation - This... this really shouldn't be an issue, but it is because the user isn't paying for the bandwidth, whether that's classic cellphone calls, Teams voice, whatever. And so its in the company's best financial interest to compress the audio as much as is tolerable. This isn't really an issue if #3 isn't a problem and the user is in a place without too much background noise, but with open-plan offices, it is a problem if there's a lot of people talking.
3. User's mic quality - So in gaming communities, people will generally tell you if your mic sucks, because you're just some random stranger. And if your mic sucks and you want people to listen to you, you will probably buy a good mic eventually, tweak the settings etc. In a business context, my experience has been that the audio quality has to be pretty bad before anyone even says anything to the speaker. And then it's up to that person to either try to get it replaced by IT or pay out of pocket for a good quality mic. And this is assuming that they're technical enough to be able to pick out a good mic to begin with or even realize that its something they can solve.
I dabble in live audio. There's a huge amount that still could be done in the audio input space. Mic quality could be greatly increased for not much money, but most manufacturers stuff the cheapest component then can find into any headset <$150. Also, there's a lot which could be done with DSP (digital signal processing) before the mic even hits the computer.
You do see high quality mics and signal conditioning on the higher end systems, usually north of $250. And even then, it feels like that's a knock-on of paying for more headphone quality.
This is all overkill, solvable with the cheapest trash <10EUR pre-covid, setting up Mumble, and using push-to-talk. With moderation this even works for a few hundred people. Otherwise it's perfectly usable for about up to two dozen people, which should suffice for most meetings?
They did get crystal clear on landline. Then we moved to cell phones where quality is still abysmal and the phone app barely works... It's aggravating to me
More like it de-materialized. Unless there was a physical line problem, regular old phone calls over copper lines ("POTS") worked well. They were circuit switched, not packet switched. You essentially had a dedicated path provisioned, end to end. Today, POTS is all but gone. Almost nobody has a real landline. Most phone calls are transported over IP. I converted my landline to VOIP almost a decade ago. It's fine.
You are right, but at least the quality was reliably poor! Today, you'll spend several minutes ask someone to fix their microphone. Eventually they realize it's not even connected to their computer.
I don't understand why audio quality is so bad on every device.
Phone calls are hit or miss whether it'll be clear or not. This happens alot with places like call centers, the whole point of its existence is to be on the phone communicating with voice and the quality is to the point where it's hard to understand.
AIUI, one of the worst cases of interoperability legacy I've ever seen. If even one thing in the pipeline is compatible with POTS ("plain old telephone service", i.e., land lines in all their 3KHz glory), the whole call degrades, and since the whole call is going to be degraded anyhow, almost everything written to handle voice calling just drops straight to the POTS lowest common denominator. Which in a digital world can be even lower than POTS due to our ability to just set a number on our lossy compression codecs with all the regard for how much money bandwidth costs and no regard for quality.
This includes hardware too, e.g., microphones that work fine in the POTS frequency regimes but don't produce high quality audio, speakers chosen just to work well in the old frequency regime, etc.
So, despite the fact I have to imagine the odds of a call hitting the actual physical POTS system approaches zero today, and that in general in 2023 a high-quality phone call wouldn't actually be that expensive, the odds of a call traversing something that lazily fell back to POTS-level standards for whatever reason is still quite high.
One could write a brief sci-fi story in a Star Trek-inspired universe in which galactic war is started because the video call to High Command in the year 2642 is still running on POTS audio quality standards and some words are fatally compromised....
To be fair to POTS, it at least made up for frequency response with near zero latency. What you describe is worst of both worlds -- latency of commodity packet switching plus bandwidth of POTS.
Personally, I'd always choose zero latency over audio fidelity in a two-way communication medium.
Audio calls are fine on FaceTime and Teams (and, I suspect, most other products). But if you and (especially) your team are still talking to your screen instead of using a real headset, then yeah, the quality is going to suck. One doesn't need use the pricey headset and mic I normally use for music production, just something that doesn't have the software DSP constantly trying to filter out background noise while still picking up your voice.
But if you're referring to cell calls, yeah, we lost a lot of quality when we ditched landlines.
Encoding is fine now. Microphones are bad though. Megacorps cheap out on providing some semi-decent headphones to the employees, that's why audio can be bad.
> I wonder if employees would personally cover the cost of a one week meetup once per month if their other option was move city and office all the time.
Nah employers paying out of pocket makes so much more sense for this arrangement, just pay the remote worker less and make it clear they need to come in for one week a month. It's basically the same thing.
One week per month in-person is probably too much. Either have short monthlies or longer for a couple times a year.
It doesn't really make sense to have the employee pay because some who are fairly close by will come in and cut the day short and those who have to pay for a flight/hotel will either shrug it off as essentially a commuting cost or will deeply resent it.
Some people make enough and are mostly fine with the mental accounting to pay for certain business expenses out of pocket but a lot of people absolutely are not--especially if it's required.
I know this is personal preference, but in the past 10 years I've used Google Chat, HipChat, Slack, and Teams, at 3 different companies, and I honestly feel they're all "good enough" to collaborate and get work done. Not once have I thought, "We'd be more productive if this screenshare were higher resolution."
Video conf indeed sucks sometimes. That's when that "video" part is mandated. Actual work meetings where people look at someone presenting something or simply listening in background and working meanwhile are just fine. As long as some lowest rung PM is not mandating turning on cameras, "video" conf is great :)
Really? Video conferencing seems dramatically better to me than 10 years ago. Zoom is pretty great and various tools for screen sharing are much more prevalent now. And the one thing I really missed about working in an office (being able to whiteboard something) is mostly solved now with an iPad + Apple Pencil.
If you work for a company where your whole immediate team are co-located it can be great. In my experience I end up spending most of my time working with people in disparate offices so I get the worst of both worlds, constant in-office video calls (and all the meeting room shuffling that goes with it).
Proprietary telepresence systems have been around for a long time and they are good enough for keeping international relations going between country's governments, so they probably would be good enough for your company. They are more expensive than you might think they should be until you get into the engineering and understand what it takes to make it seamless and reliable.
The question is, does your organization actually know the value of communication between remote parties? Companies that actually run the numbers on the value of remote collaboration can pretty easily figure out if it's worth it.
We had a Cisco conference setup for two joined conference rooms back in 2014? It was like $500K per site. It was terrible, picture quality was 1080i with bad sound, but something an exec would love cosplaying as a member of the NSC. The thing couldn't easily handle conferees using webcams etc. Got torn out when the support contract ended and converted into a traditional conference room.
Is your contention here that the technology to make this seamless and reliable experience doesn't exist, or are you agreeing with me that it's not trivial to implement?
I think that unless you're a nation-state with huge budgets, creating a seamless and reliable experience is relatively non-existent. We've tried all the major vendors for conference rooms, and they all have sharp edges that give you continual paper cuts. The same is true for tech for remote users (Teams, Slack, etc etc.). That doesn't mean they're not good enough, but they definitely still suck.
I agree it sucked Pre-Covid, but now it just works. We mostly use Teams. We are RTO 2 days a week, but most work meetings are still in Teams. Demoing is way easier because you can simply share your screen rather than carrying your laptop to the various meeting rooms.
Why do so many want to video conf specifically when audio is all you need? Hell, of these, text chat is all you need and is more practical most of the time.
Because audio-only or, worse text-only, is throwing away >90% of the bandwidth of human-to-human communication. Studies show that relatively little of in-person communication is the words themselves. I don't think emojis solve this.
But it makes me wonder how people have time to invent all this shit when they need to move fast. Learning a new stack every 2 years doesn't seem like moving as fast as you could.