Interested in the outcome of both of these cases, but also in what went in to building the site.
It feels agent-led, but in a good way.
@laurex are you the creator? Can you tell us some more about how it was built?
187 karma · joined September 20, 2013
@laurex are you the creator? Can you tell us some more about how it was built?
"Except for vaccination for adults aged ≥65 years, ACIP makes no preferential recommendation for a specific vaccine when more than one licensed and recommended vaccine is available. Among adults aged ≥65 years any one of the following higher dose or adjuvanted influenza vaccines is preferentially recommended: HD-IIV3, RIV3, or aIIV3. If none of these three vaccines is available at an opportunity for vaccine administration, any other available age-appropriate influenza vaccine should be used (4,5)"
There's a key standard for this: Digital Public Goods, which you can read lots more about here:
https://digitalpublicgoods.net/
https://unicef.github.io/publicgoods-toolkit/
https://digitalprinciples.org/#principles
If this sounds like fun to you, here's a couple of my favourite Digital Public Goods that have great communities and make real impact:
https://github.com/primeroIMS/primero
'I'm sorry your honor, I have to withdraw from this matter because my client, who is a bit dim, has just told me at the lunch break that he totally did the thing, and also that he's going to lie his ass off about it when he gets up in the witness box in just a minute'
But the next time you hear a lawyer withdraw because they're 'forensically embarrassed', you'll know what happened
I'd sat through 'five whys' style postmortems before, but it was reading air safety investigation reports that finally got me to understand it and make it a useful part of how we get better at our jobs.
By comparison, the way we're investigating and responding to self-driving safety incidents still seems very primitive. Why is that?
And isn't that as big a story? That in 2023, in the richest country in the world, there's no publicly searchable index of the laws of the country?
From the infamous irony quotes 'balance sheet': https://d1e00ek4ebabms.cloudfront.net/production/7ab64a3b-6c...
Liablities: $8.8bn of customer deposits
* $6.6bn-ish in USD or USDT
* $2.2bn in crypto
* I'm assuming the 'poorly labelled $8bn' is included in this amount, and you don't add it
Assets: $2.5bn?
* $1.5bn of liquid assets.
* $1bn-ish? of illiquid assets.
* Most of these numbers are just made up, but let's finger-in-the-wind them at $1bn ($500m at least is in 'locked USDT')
Now we find out they have another magical pool of money, the $3.2bn that they'd taken out of FTX and put in their pockets.
So the position looks really different: They had $8.8bn on one side of the ledger, and $5.7bn on the other side - After the massive bank run which collapsed the price of their shitcoins.
It appears they had the money to pay out most of their USD / USDT depositors at $0.85 on the dollar, which is obviously still 'insolvent', but possibly enough to survive a bank run (almost) all the way down to the ground (If you suspend redemptions of your crypto and persuade 15% of your USD depositors not to pull their money out)
The final question is: Who did they owe the liabilities to? If some of those 'deposits' were actually owed to other FTX-associated entities, it seems even more survivable, and would probably have preserved some of the value of their shitcoins
When I asked him why he'd gone so all-in on vegetarianism, he said "Daddy, animals are my friends. And you don't eat your friends"
The little bastard derived the whole thing from first principles, with no input from his parents.
I've tried to use the vegetable-based mince alternatives in pasta sauce, and we're always really unsatisfied with the results. He much prefers lentils / tofu / tempeh and frankly so do we.
We're in the target demographic for these products, and I can tell you: THe problem isn't the price, the branding or the level of ethical awareness. It's the product! It's a lousy substitute for the meat-based alternative, and it doesn't even stack up well against the raw ingredients they're derived from, which have the advantages of being cheaper and tasting better.
That's really sad news. The crew I interacted with (Alex, Omari, Tobias, others) were top-flight engineers who also really understood the problem space and how to navigate the internal complexities of Meta. Hope the laid off / repositioned staff are OK, and if they're looking for new opportunities my email's in my profile.
For comparison:
Meta / Facebook:
* React
* PyTorch
* Fasttext
* Their unsung-but-indispensible-for-humanitarian-orgs Data for Good program
Google:
* Flutter
* Go
* Kubernetes
* Tensorflow
* BERT / Large language models generally
* Mapreduce / Hadoop
The much-maligned Uber:
* H3 Geospatial indexing
* Kepler.gl
* Manifold
Even AirBNB!
* Airflow
* Superset
I use most of these products (or their descendants) every day in my work. When I think of Twitter, all I think of is how they arbitrarily shut off humanitarian access to their APIs (and earlier, the Twitter firehose). There's not a single Twitter-supported open product that I use.
Am I the only one? Is there a great Twitter-supported project that I missed?
To rub it in - If Facebook & Mastodon were the only two social networks to 'switch' to, FB captured 99.8% of the 'switchers'. If 24 Million new users had signed up for Mastodon, that would be a story. That only 0.2% of them chose Mastadon shows how bad the open, federated alternative really is.
Don't delay the decision - You've had performance conversations with them already, so this won't come as a surprise.
Tell them they're on leave immediately, and then pay them whatever they're owed in terms of leave, set their effective termination date for just after their vesting date.
Talk to the team, explain that it didn't work out, but that you're not the kind of founding team who will screw someone out of their equity weeks or months before their cliff.
The team will respect you doubly : For getting the under-performer out of their team, and for treating them fairly. This may sting a bit, it's your company and you're killing yourself to try and make it worth something.
Consider that the cost of learning how to be a better boss and not having made the hard decision earlier.
If you go the other way, you risk killing your velocity and losing the people you need.
I get the occasional tab freeze / crash, and it has some glitches with Google Docs / Sheets, but it just feels so much faster and uses much less RAM too
You have an audacious, but specific goal. That's great!
You have at least one person who cares about the outcome of your campaign (you). That's good, you'll need more
1. You need a power map. https://commonslibrary.org/power-mapping/ is a good resource to get you started. Get together some mates, preferably ones who've interacted with the political process before (They'll be more likely to know the players) and start mapping. Once you've got a good power map, that's a good time to reflect: Does this campaign still seem achievable? Is there too much power in the top-left quadrant, and not enough in the top right / middle? Movements that lack a good 'center' of persuadable+influential actors generally fizzle out, so you might want to consider tailoring your campaign goal.
2. You need a theory of change. How do we get from here to there? https://www.thechangeagency.org/?s=theory+of+change is a good start. Don't lock in to a particular tactic too readily, and be prepared to iterate
3. You need power. This both informs and influences your theory of change. Is your power electoral, financial, social?
4. You need a logframe. https://en.m.wikipedia.org/wiki/Logical_framework_approach How will you know if your campaign is working? What intermediate metrics of success can you identify that are both impactful and measurable?
There's a mountain of campaign theory and literature to read, but that's enough to get you started assessing if your campaign is a goer at all. Hope it helps
And she said 'Well, that seems easy. I'll just talk lower'
This is the kind of thing 'growth hacker' types would normally idolise, rather than ridicule.
I think the voice thing has been conflated with the fact that she was running a huge scam, which we've decided is generally ok when you're making a food delivery app, but is definitely not ok when you're testing people for HIV.
Fake-wood plastic floors
Fake-wood MDF cabinetry, with superfluous detailing
Plastic benchtops
Horrible overhead lighting layout.
The 'LCD television-as-wall-divider' is the cherry on top of this nightmare.
The problem isn't simply that they've chosen cheap, poor-substitute materials, the whole thing has been thrown together with no sense of design or cohesion.
Good design costs almost nothing - Contrast it with this Australian modular / pre-fab: https://www.dwell.com/article/wattle-bank-tiny-shipping-cont...
It also uses inexpensive materials (Untreated plywood for the interiors, shipping containers for the structure), but it's thoughtfully designed and quite beautiful.
He creates a classifier with 5 label,s and gets 50% accuracy on the cross-val (This may be ok, may be terrible, we don't know), but then only uses the labels for one of those classes. What's the precision/recall on that class?
Also, he's taken a multi-class classification approach to what is essentially a regression task. He's trying to predict the # of upvotes a post will get, and there are approaches that'll do that, but this isn't one of them.
It's fun to see NLP applied to these sort of problems, but this isn't a good example of how to do it well.
It'd be good to have some better incentives for management to resource and schedule properly (time and a half @ 40hrs, double time @ 48, quadruple time @ 60 would seem an obvious economic stick), but if employees are getting paid fairly I think this is ok. Hope the game is great, and that everyone who is putting in the extra work feels tonnes of pride of authorship in this game. I'm looking forward to it.
In the case of Facebook, I imagine there are people who read the headline + snippet on Facebook, look at the photo and then don't click the linked article, which seems more like actual infringement of the copyright. It's not the same as taking the entire newspaper and replacing all the ads with your own, but it's a bit like that, enough that I can see what the intention of the law was. It's all complicated by the fact that the news organisations currently facilitate the process, and obviously at some point they thought it was worth it