jelling

You might like Cog. It solves these problems for ML projects in specific.

Guidance from Microsoft or Jsonformer might help. Haven’t tried with vertex but it’s a common problem across LLMs that both projects fix.

The problem with Reddit as a business is that they have the least profitable users of any social platform. So one would have to build a backend, incentive the app makers to switch, get advertisers and do it all more cost efficiently than Reddit does it.

That’s a small needle to thread.

CivitAI.com is down and Hugging Face is having 502s for some operations. Perhaps there is a larger issue?

Re: hiring contractors via an intermediary, the civil service unions pay scales make it nearly impossible to hire skilled specialists at market rates.

Also, it is very difficult to lay-off public employees if their role becomes unnecessary in the future.

Hiring indirectly via a contracted intermediary largely solves both of these problems.

Deep Floyd doesn't allow commercial usage, such as a game studio using it.

It's "Criti-Hype", basically using FUD to get media attention to lower CAC.

Source: https://sts-news.medium.com/youre-doing-it-wrong-notes-on-cr...

The commentators point was that YouTube may not be as strong of a stand-alone company as the article believes if the COGS is inaccurate because YT gets tons of free resources via Google.

I suspect the primary driver for 1 - 2 week retreats is that is the average Western vacation length and the mindful companies are maximizing revenue relative to CAC.

Formal Buddhist training - such as to be a teacher within the Kadampa tradition - can be done once a week over a course of years. Other traditions training requires full-time reading training over years.

I wouldn’t recommend someone try to master an instrument in 7 - 14 days as learning takes time for integration and people progress at different rates. So I’m skeptical of why anyone would suggest that for attempting to master the mind.

I bought one and regretted it as there is a strict no return policy. My purpose was possibly using it for AR development but I found the experience not much better than Google Day Dream, as far as I can remember, and possibly worse because the hard plastic molding made it harder to position your phone. Also I missed the Daydream remote.

On the upside it disabused me of the idea that AR was any closer to ready at a relatively low consumer price point.

Per typical AR product, the preview videos were wildly better quality than the actual experience.

I used to work in restaurants and have enormous respect and sympathy for anyone that tries to make a go of it. But the featured protagonist didn’t even have enough cash to last more than 30 days when even the article says that Cloud Kitchen promotes the idea that they could be profitable in 6 months.

Said person then went to the press and tried to start a class action suit. I think they need to look in the mirror and recognize how woefully unprepared and likely unqualified they were to run a restaurant.

And no, Cloud Kitchen doesn’t want to have people fail out that quick so they can “get the deposit.” That’s a terrible LTV to CAC ratio, and like the article says, they can’t even fill the place as it is.

The difference is that winforms basically had one way to do everything which expedited development because it skipped all of the infinite number of stupid small decisions and libraries that must be cobbled together in a typical SPA.

That and a lot of engineers are genuinely excited to build their own workflow engine - whether they call it that or not - because it’s complicated and it feels like they just discovered a brand new and powerful abstraction.

Tears follow when the team either doesn’t account for all the edge cases or doesn’t have the resources to address them.

Paid news letters are the “publish or perish” of the non-academic world.

Any plans to expand to Discord? Game developers and others working with non-enterprise users have the same problem.

Your comment is a perfect example of what I was trying to head off: people that...

- don't live here (it appears you live in Europe?) - don't know anything about the local politics - don't want to listen when anyone tries to tell them - don't know much about the justice system (including the difference between jail vs prison) - use ham-fisted statistics like national incarceration rates but... - don't account for enormous variables (hint: guns), or... - weighting of prior sentences (it's a rolling average obviously)

But surprise! they want to squeeze every story into a large conversation so they can take the moral high ground against straw man arguments no one was making. Like this one:

> Why would you not want more of your citizens to live fuller, more accomplished lives rather than rotting in shitty prisons?

You took time out of your day to write that comment to some that told you they supported carceral reform. Heck, you even quoted me saying that I wanted carceral reform and thought "Nah, this guy needs my input about on whether we should let people rot in prison."

Here's a policy recommendation: don't run your mouth on topics you don't know much about. Greetings from New York.

New Yorker here. If this happens, great, but ysk that building those replacements jails is a knock-down drag out fight on the local level. Even renovating/replacing an existing jail in my neighborhood is being opposed by people that literally want to get rid of jails altogether. Literally, we can't close the terrible, poorly run jail (Rikers) because some people don't want jails at all.

So take this article with an enormous grain of salt. And ymmv, but articles that simultaneously play to green energy projects (good thing) and carceral reform (good thing, more complicated in reality) are likely high-minded click-bait.

> Stripe is a hugely successful company and they have no urgent material need to let people go

Indirectly, what you are suggesting is that the company string people along for as long as possible and give them busy work. So people would get the money, but not their time.

What Stripe did with their severance package is give employees both the money and their time back. Few people would likely prefer still having to go to a pointless job.

(And of course Stripe had an urgent need to let people go, they wouldn't have the money to pay such significant, if any, severance.)

There is no "zero costs" way to operate an economy. If we increase the long-term responsibilities of a company to their employees - as opposed to giving the same amount of safety net via public means - it will have a significant impact on willingness to hire.

Stripe's severance package, which is as generous or more so than the most advanced democratic socialist countries, is about as good as one can hope or should hope to get from a company.

If longer benefits are desired, the voters of California would need to come together on that and figure out how to finance it. But conflating the LA homeless / drug crisis with the 14-week severance packages for high skilled workers doesn't add up.

(Side notes: the perception of job loss in Japan is drastically different than in the U.S. and the Bank of Japan has been lending money at near zero rates for decades. The result has been a plethora of zombie companies.)

This website is pretty bad in terms of pitching the product. On a Macbook Pro, there is a giant black bar with nothing but the headline, which is then repeated in the video. The video itself tells me nothing about how it can improve an enterprise. And The macro fly-by of the device itself makes sense for a B2C product much mote than it does for a B2B product. Just weak across the board - feels like it might have been done internally even - and that's before we talk about how they've raised $3.5 BILLION dollars.

In another life, I managed a team working on Google Glass and tbh I'm not surprised that they still don't have more compelling use cases. Hands-on training is a relatively niche market. And the medical example - separating conjoined twins - seems like it could be handled just as well or better via VR or just a standard display. The dirty secret of the space is that enterprises want to de-skill labor wherever possible to save costs. But if a step in the value creation process can be truly de-skilled it likely can be automated, and if it can't, human beings learn well enough that they don't need (or want) the goggles strapped to their face for long.

I managed a team building for Google Glass. These high complexity labor use cases were the promise then as now. But the quiet part is that it’s to de-skill these positions as much as possible. Problem is that in the meanwhile they want highly skilled people to put up with basically nanny-ware strapped to their face.

Re: lowering interest rates, China has the largest foreign currency reserves in the world at over 3 trillion USD. So where most sovereign currency issuers can’t print money without reducing the value of their currency - or increasing interest rates to compensate - China can defend the value of their currency by purchasing with it foreign reserves.

I’m not an economist so I can’t tell you how far they can go with this. But they have an atypical margin for error correction re: interest rates / money creation.

Yes. But as someone that's been meaning to switch off of Evernote - never has a project bloated their own software with so many unasked for features / desperate attempts to grow valuation - it's a non-trivial project. One needs to have a high degree of trust in the import tool and new repository.

This is also why business that rely on Google ads suck.

This is incorrect on a political and economic level. Fwiw I live in lower Manhattan, so I’m not saying this because I benefit from suburban subsidies. But if the country as whole subsidizes roads and services for a suburb of net positive income tax payers with otherwise low usage of welfare services, there is an economic logic to that. And obviously it gets votes.

As much as I like living in a dense area and riding my bike, the rise of the YouTube educated urban planner has led to a lot conclusions that aren’t particularly sound. Doing a flaw P&L on a small town and deciding, in so many words, that it shouldn’t exist is peak smugness.

Combining the Discovery and HBO brands is like taking two random people off the street and making them conjoined twins.

Now explain why they put carpet in a train.

For some hard data, in San Francisco and NYC, the mass transit agencies are starting to plan for never getting back to full ridership again. And by plan, I mean starting to ask for federal funding to keep the systems alive (which I think they should) because the fare revenue is never going to return to pre-pandemic levels but the physical plant remains just as expensive to operate.

Keep in mind: these are two of the densest cities in the U.S., where people typically have less space for a home office and commutes are shorter vs an exurb type city. WFH is here to stay in my view, especially for employers in competitive recruiting spaces.

> Code quality remains an abstract concept that fails to get traction at the business level. Consequently, software companies keep trading code quality for time-to-market and new features. The resulting technical debt is estimated to waste up to 42% of developers’ time

Note: this paper was written by an employee of Codescene, the tool used in the paper to analyze code quality.

A codebase that is brittle - meaning things break easily, testing is low, or where maintenance or new dev is difficult - is a genuine business problem. The solution to that is engineering managers that can make a persuasive case to non-technical people for why something needs to be done a different way, or failing that, will black box that entirely and add the needed maintenance to an existing business initiative. Put another way, play corporate politics like regular politics and tack what you need onto a bill that must pass. Showing an executive a report from code analysis tool is unlikely to work.

But as a founder and an active software engineer, far too often "technical debt" is used by programmers to advanced their personal preferences, which may not be aligned with the businesses in any way. A few highlights of things I've been told are technical debt in recent years: not using micro-services (when we had no need for the added complexity or scaling features), using promises instead of observables (in a non RX codebase), using a mono repo (again, zero business problems caused), using any number of 3rd party libraries (vs the developers desire to build from scratch), using Github instead of GitLab (or the reverse), etc...

> Or maybe the real problem is the excess emphasis on finding a drug in specific that can be commercially monetized > ... > Amyloid and another protein (tau) accumulate in the brain due to sleep deprivation. Helping people get a good night's sleep -- without drugs -- would likely help, if only to stave it off in those at risk.

People are dying right now. Giving them better sleep habits won't save them or meaningfully extend their lives. People are suffering in ways that I don't think you understand and I hope you never do.

I feel similar but just want to add that letting people do a ~20 hour probation week could be useful for both parties here. Meaning they can keep their same job, but you're going to pay them to do 1-2 weeks at 20 hours per week. This doesn't sound like a sensitive IP situation, either.

Also, if you can't get someone up and delivering value in 20 hours that's good data about your internal setup. My goal is a minor commit on day 1. And if you can't spec out value that could be delivered in that time frame you probably have weak requirements.