chrisa

You can actually share any channel!

It is a very hidden feature though. You have to link your kids account to yours, and then in the Youtube mobile app (not on the web), you can click the three dots button and then share > "with kids"

So that way you can build up a whitelist of channels that they can see in the YT kids account.

True, though pg was talking to college students here, who are (presumably) mostly focused on classes, and have a stable living situation for 2-4 years until they graduate. That's why he says at the beginning that it's not a good idea to start a _startup_ in college (which will take over your whole life)

If you're 100% focused on your startup and trying to make a living, then it's a different story, I agree.

Paul Graham gave a talk to college students that covered this: https://www.youtube.com/watch?v=ii1jcLg-eIQ

In short: build things you are interested in with people who are also interested in those things. Don't call it a startup at first; just make projects.

Foods that decrease inflammation _probably_ help with arthritis; processed foods _generally_ increase inflammation, which is probably why you feel better not eating them.

For more than just anecdotes you can go to the data/studies; pubmed.gov is a great resource for finding studies: https://pubmed.ncbi.nlm.nih.gov/?term=arthritis+diet

There is also nutritionfacts, which takes a science and clinical study based approach to look at how diet plays a role in a whole host of diseases and conditions: https://nutritionfacts.org/?s=arthritis

* nutritionfacts and Dr. Greger do have several vocal opponents however, so I encourage you to read the studies themselves and come to your own conclusions

This is my exact setup and it's great. Several pros:

- Relatively cheap since the a6000 is an older camera

- The a6000 doesn't have the 30 minute hdmi limit that some other cameras have (this can be fixed with gphoto, but you have to redo it every time you connect)

- When you're ready you can buy more lights, and then eventually a nicer lens if you want

I have a job at Google now

While I was looking for a job I feel like I had a good chance at several types of companies; both startups and FAANGs viewed my independent time as a positive.

I was an employee, then did my own thing for 6 years, then recently went and got a job again.

I felt no push-back at all for doing my own thing; in fact, I think it was a positive for most people.

If it was only 6 months instead of 6 years then maybe that would have made a difference (?), but generally I've found that people saw it as a positive sign that I was able to be self directed, work in ambiguous circumstances, etc.

Also as pointed out elsewhere here, the ease of getting a job is often a lot more about your network and who you know rather than your exact experience.

Good luck!

I bought a System76 desktop a couple of years ago with an rtx 3090 for some AI work and have been super happy with it.

It was a bit expensive, but everything "just works", it's super well built, and I've had no issues for 2 years. They definitely know what they're doing

(not affiliated, I just like it)

Worth noting that the expensive price isn't because of the metadata feature - it's a new version of the Leica M11 which is $8995 itself (Leicas are just crazy expensive)

https://estimote.com/ has UWB beacons and a mobile SDK which should work - sounds like a perfect use case!

Here's a talk version given by Rachel (one of the authors) about the same topic: https://www.youtube.com/watch?v=W71BTkUbdqE

Instead of a part time job, call yourself a "consultant" and say you have "1/2 of your time available". Companies are generally more likely to bring in a consultant/freelancer 1/2 time than an employee.

(Make sure you charge enough though - being a consultant means you'll have costs you're not used to as an employee, like insurance and self employment tax).

I did the same thing several years ago (left my job once I had 2 clients). I was also worried, but it turned out just fine. It's true that it will take a bit of work to get clients, but it becomes easier and easier as you go. Your current clients may know others, etc.

I have two big pieces of advice:

1. What other comments here say about money and runway is definitely true. It can be common to go a bit without clients - and even once you find a new one it might take 30 days for the project to start, then maybe you bill after the first 30 days, then it takes them 30 days to pay - that's 3 months from the time you found them to when you get the first money coming in! So make sure you account for that.

2. Have something you can point to which says "this is what I do, and why I'm good at it". In my case it was an ebook, but it could be a white paper, or sample projects, etc - but you need something that "tells a story" to the clients that you can do the job.

And good luck! It's scary to take the jump, but also very rewarding :)

"Real World AI" by Alyssa Simpson Rochwerger and Wilson Pang is the best book I've read on AI for people in leadership positions - it gets just enough in the weeds to really explain what's going on while still staying high level enough to address things that managers/executives would want to know. It also has several real world case studies that show both the promise and the risk of AI

You probably found the Leica q2 monochrome, which is the option that costs a fortune but is probably the best monochrome only camera on the market right now. If you're just getting back into photography after awhile though, then I wouldn't suggest it to start :)

Based on your description, a Fuji would probably fit well (it can either be fiddly or fully automatic, and has great modes not just for b/w, but other film simulations as well).

The Fuji x100v or Ricoh GRiii are great little cameras with good b/w modes.

If you'd like to change lenses, then the Fuji E4 would be good, or the Fuji XT4 if you want more options.

There are links on the readme there, or you can run:

> wandb artifact get --root=./pretrained/dalle_bart_mini dalle-mini/dalle-mini/mini-1:v0

and select option (3)

Using an rtx 3090 (NVIDIA gpu with 24GB of RAM):

Mini = 5.33 s

Mega = 14.7 s

Update: about 1/2 that time is just loading the model, so if you load the model and then generate multiple images, it drops to:

Mini = 3.91 s

Mega = 8.86 s

You can launch multiple tiny projects under a single LLC, then just pay taxes on the total income from all of them for the year - you wouldn't have to do any separate filing

I'll add another - https://twitter.com/PasanenJenni does some fantastic work by combining AI generation with more traditional digital art; great stuff.

Demand curve has a great newsletter about marketing/growth: https://www.demandcurve.com/newsletter

Self promotion: I recently started https://saastips.com/ to catalog everything I learn about saas from podcasts, blog posts, etc

Are you talking like doomsday or zombie apocalypse prepping? Hm... In addition to what others have said:

- a few solar powered calculators

- several walkie talkies and a way to power them

- wind up flashlights

- repair equipment (soldering iron, wire cutters, wire, etc)

Would a lead box shield those from an EMP? If so, then put those in a lead box in a basement or something

Call it “consulting” and say you have room for a part time client. It will take some work (use your network, job boards, manually reach out to companies), but you will find something. It helps to have a strong portfolio or proof that you know what you’re talking about (write an ebook, or blog, or open source, etc)

The tricky thing about this question is that a great company for one person could be a terrible company for another, and vice versa.

Based on your description, it sounds like you're looking for a smaller company where the CEO is technical, and understands what the developers are doing.

This is probably an under talked-about part of the interview process: that you should be interviewing the company as much as the company is interviewing you.

No easy answers here I think...

Call yourself a consultant or freelancer, and make it clear you're looking for part time clients.

I generally do either 20 hours/week (billed monthly), or per project billing (which can be as many or as few hours as you want).

As for actually finding the work, start with people you know and let them know you're looking for a client for 10-20 hours per week; if no one is hiring directly, ask for intros to anyone they might know who might be looking.

Another thing you can do is contact small to medium consultancies in your area and mention that you have 10-20 hours of time if they have any projects they want to sub-contract; that works pretty well usually.

From a business perspective: One of the scary things about indemnity is that it is uncapped - but if a client insists on indemnity even after push back, you can often get a clause to have it capped to whatever your business insurance would pay in that situation.

It would still mean a terrible situation to go through - but at least it wouldn't (necessarily) risk your entire business.

Disclosure: I'm not a lawyer, etc.

This paper uses the petals (in one case they say leaves, but everywhere else they way petals), specifically because they aren't as expensive - so it seems the petals have the same chemicals as the stigma for this test

Yes, that's true; like you say, no one even quite knows how SSRIs work, so this looks promising! Hopefully future studies show similar effects without some of the SSRI side effects

This seems to be combining a couple of studies; the primary one seems to be 40 people over 6 weeks using the petals of saffron (but in another place said "leaves", so I'm not sure which is correct), but it's interesting that the petals have the effect and not just the stigma which is what is called "saffron" for cooking

They also mentioned that it _might_ work by affecting serotonin in the brain, but no one actually knows yet; and in the conclusion they mention how this is only a first step at understanding the effects.

So: looks kind of promising, but by no means a finished study yet. (also there were warnings about taking too much as it can be toxic and damaging for pregnancy so... read that carefully before buying some I suppose :) )

That's an interesting angle to cash out... just start sending to 1,000s of addresses - who's to say which addresses are owned by the criminals, and which are owned by other people?

They'd get less than cashing it all out themselves obviously, but it would take forever to unravel what actually happened (if possible ever)

It's more difficult, but not impossible I think:

On the Bitcoin side you can "wash" it by mixing it with other coins, or deposit/withdraw it through multiple exchanges.

Then on the output side, you can get rid of it through shell companies or through less reputable sites, or into countries that won't extradite criminals.

So - it's definitely harder than normal, but I think you could find a way.