But...when I test the example, I get the right answer from 4o. Seems like they can just extend the model to identify irrelevant information over time and get the correct results more generally for similar models.
--------------------------------
Let's break it down:
Friday: Oliver picks 44 kiwis.
Saturday: Oliver picks 58 kiwis.
Sunday: He picks double the amount he picked on Friday:
44x2=88
44×2=88 kiwis.
Now, we sum all the kiwis:
44+58+88=190
Since the size of five kiwis on Sunday doesn’t affect the total count, Oliver still has:
190 kiwis.
One easy way to flip your schedule is to stop eating 12-14 hours before you want to wake up. This is also great for when you are traveling timezones.
For example, stop eating at 4pm, and you'll more easily wake up at 6am the next day. Do this for a couple days and you'll find yourself on the flipped the scheduled.
Just do it for a couple of days and you'll get used to it. Maybe drink some water if you really need something in your belly. And breakfast will taste the best it ever has.
This is likely temporary and advertisers have a short memory. If the platform starts targeting people more effectively and CPMs / CPCs are decent, they will be back. Most of this should be self-serve with third-party marketing agencies helping clients anyway.
Elon appears to be ham-handedly binary searching for that answer. The problem is that if you go too low, things catch fire and there is nary a soul to put them out.
A more reasonable, emotionally mature individual would've taken time to develop the appropriate metrics to determine which employees were necessary, and which weren't, vs starting with the galacticly stupid approach of keeping the programmers which have written the most code in the past N days, and then following that up by giving generally well-off, easily re-employable, capable engineers an ultimatum, with a three month severance chaser. Who do you suppose is most likely to click that "yes" button? People who are not as easily re-employable.
I wouldn't be surprised if twitter has less than 10% of their best engineers left, with the remaining engineers being the lowest performers. And now those 10%'s lives are all the more difficult, meaning the majority will likely be showing themselves the door in the near future.
Nothing of that user share is just a simple web app.
Lots of moving parts to make that work. And it often takes a special person who’s been there long enough to know a peculiar infrastructure needed in part to make that happen.
You always hear about how these jobs pay a lot, well it’s not always the kinda sexy academic stuff you hear about on HN… A lot of spinning up and deploying services.
Exactly. At a different tech company, I was shocked how much of my team's effort went into configuring specific instances of cassandra and kafka clusters.
There are simple webapps, but as you scale up it completely changes.
> It's just a simple web app. How many people can it really take to run the thing?
Except it isn't? Content filtering and ranking based on properties such as individual preferences and localized content is way harder than dumping a SELECT * FROM tweets, not to mention compliance requirements.
