maytc

Is it terrible at the content or the delivery of the content? I find it to be bad at the prior but rather useful for the latter.

If you asked chatgpt to generate a document, its generates one that reads well but has terrible content. Eg. Super broad or just plain contradictory. But given some not so well written piece of writing, it can clean that up fairly well.

Surprisingly, I found the flush makes the phone much easier to pick up from a flat surface like a table.

You can ask it to rewrite the given prose and make it funnier. If the given text is already kinda funny it should spice it up. If it doesnt have much material, gpt may likely hallucinate some material to get it to be funny

I've been using it as an aid in my writing process. Essentially, I can now pour my thoughts onto paper, complete with errors and poor word choices. With just a click, GPT transforms it into a presentable version.

Basically I created a small app to streamline the workflow. https://henshu.ai

Coinbase Japan didn't last 1.5 years. They officially entered Japan around August 2021.

https://www.coinbase.com/blog/coinbase-launches-in-japan

I made a boardgame in 2021 and got it fully funded on Kickstarter. It's a faster version of UNO where players can play out of turn.

https://www.swiftwave.io/

My other favorites are Pandemic Legacy (Season 0 came out in 2020) and The Crew (2019)

One simple method: You start the search with a 50%/50% split, in this case half men and half women, then hire the best candidate from that cohort

Love how they are selling NFTs of the "unreal estate" on the next tab.

My previous company did this and it worked well enough! We used deployment groups to control traffic and specialize the monolith to serve certain part of the service

Im getting off my SE and getting the newer 14 because of battery life too. Weight and size, the SE was perfect.

Is it me or has the marketing leaders at Apple have taken over the presentations?

GPT-3 also understand code to some degree so it can run the code as instructed.

https://mayt.substack.com/p/gpt-3-can-run-code

Iirc Celsius was margin called over the weekend and nothing about that was covered here… wow.

https://www.google.com/amp/s/wccftech.com/the-much-battered-...

If Elon was on the board, then he would be even more liable for what he tweets. And i think Elon wants more liberty to tweet without the sec on his back even more.

Sounds like he has the best of both worlds holding the stocks and staying out of the board.

Microservices for a small application developed by a handful of developers

And Rivian is paying Scaringe for his services at 1.3MM a year. Crazy for a company with no reported revenue

Jeff Horing via Insights Partners holds 42.7% of the company options... Wow

Given tether makes up more than ~60% of the bitcoin liquidity I would argue otherwise...

https://coinlib.io/coin/BTC/Bitcoin

From my experience working remote, there are two intertwining issues. Should geography factor into an employee's pay? And how much should pay differ in different geography?

My answer to the first is yes. The second however is a bit more nuanced.

The first case is when employers find the best and brightest talent from all around the world and expect these employees to work together and deliver similar value. In this case I think globally fixed base salary * a cost of living index multiplier capped at 10-15% difference makes the most sense. You don't want the pay discrepancy here to be too large as these people work together on the same level. For example, using local market rates, a Bay Area employee will likely earn double someone in the EU for the same role.

The second case is when employers want to outsource work to a cheaper labor market. In this case, expectation is that the workers will not be of equal skill and/or the work can be done with little training. Another way to look at this, the relationship is more hierarchal. The HQ is managing the remote worker for work for example. In this case I think a competitive pay relative to the location's market rate make sense.

And deployment scripts that are unit testable. That's been years in the making.

For the which is heavier question, does it always pick the latter option?

That’s how nix guarantees reproducibility. The compiled artifacts can be cached but that didnt quite work out of the box

> Why can't we compress encrypted messages?

Not sure if eli5, but you can compress that message. Counter example: Huffman encoding that message.

Not sure where in history people started to believe that micro-service architecture is a simpler and easier-to-operate architecture. Each service needs the same operation overhead as one monolith.

The big blocker most monolith faces as the application gets bigger and is deployed into more and more machines is that _releases becomes a bottleneck_. Scaling monolith's applications are difficult because partial rollout is usually not possible as "services" are often tightly coupled.

Micro-service architecture forces services behind a set of APIs. While the APIs may have breaking changes, each can be independently deployed. In other words, teams can do releases at their own pace.

The main cost-benefit analysis here is how important is independent releases vs the cost of operational overhead?

Looks like the log log size's an upside down W for Windows.

Great call! Thanks!

The author confuses mobile web for mobile app traffic which isn't accounted for. One is the traffic from users using facebook.com via a browser on their phone versus using the installed application. Maybe the drop is real or Facebook has successfully converted more users to the Facebook app, Messenger app, etc.

Ridiculous...

Took over the universe with no cheats. Great game!