kshacker

I am technically not saying that. I am taking off from the parent comment and speculating.

Their data from September is leaked via IPO preview, and October is not yet complete, so how can someone even project that?

// I am a techie, not a finance guy, so apart from public chatter, no inside knowledge.

And I do feel their IPO is being rushed before anyone has enough information to make up their mind. The VCs or private equity have given them a lot of money and want to cash out. When the revenue does slow down, any investor should pause to take a look at whether they can right the ship, or has the landscape itself changed.

That's why Anthropic are rushing the IPO. You can explain a single quarter slowdown, you can not explain a multi quarter one.

One of them, sure. 8 of them?

But they launch a lot of satellites to build their telecom, no?

I have some sorta skin condition that I can not stand either the heated seats and the heated steering wheel.

But my wife ... in winters, the first thing on entering the car is to activate the heatings seats. Some people like them. And yes I paid for the subscription (or a package that cost me a bunch but gave me everything all inclusive).

Someone's got to get the vulnerabilities patched before AI (sorry, SI) does a coordinated hit job. I like this one at a time thing than a complete dystopia. By next year, half of these will be patched, and people writing more secure software .... sorry one too many jokes :)

And like any complex software, the last 20% takes 80-90% of the time.

I must be one of the few who completed the interview. I was hesitant with some questions but still gave direct feedback and asked to make my feedback public. Not sure if they will do it. And just like LLM does the interview did pick up some nuances from my answer and played them back for me to confirm. "Oh so you are saying not only this but that, is that right?

Whether they use it or not, whether they use it to manipulate us or whatever, it is undeniable that AI is the biggest change happening right now, and if there is an opportunity to nudge it, why not.

Edit. And no, I was not asked to login. They likely had the login. I just clicked the survey link from the terminal claude, it opened in a browser and off to the races.

I did that last week for my son. Had photo from Walgreens though but rest of the process was quite amazing. This was before the announcement and I may have used state.gov or something ... maybe they had this going on for a while.

Still, much better than how it used to be. You could not do expedited processing but I guess it will come eventually.

Also did not need a login at all which seemed surprising and I did check whether I had gone to a phishing site.

Whether it be corporates, Eiffel Tower management, or religious groups, I guess the new norm is - better to ask for forgiveness.

I am Indian origin so full disclosure done. I saw a picture of the group's 93 year old head in ascetic robes, the one who does not see women, have a picture below Eiffel Tower for memories. Sir, what part of asceticism do you teach?

Nature decays, but latinum lasts forever.

- Rule of Acquisition 102

What triggered me to ask was because I can see myself writing that. Not the content but the style. In the old days, I will spend time finessing an argument and feel happy about it. But with the pace nowadays plus the AI, it is a different world, so do not know.

... or frustrated. "I have my quota of jiras to meet, and this person is taking up too much of my time when it was such a simple issue which no one complains about". Not a good customer service mindset, but frustration happens.

What made you make that assessment?

I am curious. When you put this money in kids names, do they (or we) get taxed on this? If we save on taxes, then for a person (or their child) with high income and high state taxes, the benefit of the 3% may be larger than it appears to be. Of course, up to a limit.

Lets just break the lynx users for once, and see who complains. Change it to 81. Small enough to be invisible to most users, but when it breaks, someone will come up and shout. Better than going to 100, 132, 140 or 280.

I don't post often or ever so don't remember, you can't change the title to be more reflective of what it is?

A shorter version may be

> A Vermont pharmacy chain's AI implementation led to delays, incorrect information and privacy concerns.

So at the end of it, if we win enough lawsuits to demarcate some knowledge out of bounds, sufficient enough to make a difference, I wonder how that will affect the AI. Make it dumber because it does not have that data, it make it smarter since it will need to reason better with smaller knowledge base.

> (sorry for mentioning the TC, but there is a reason for it ....

sorry to nitpick on this, but the story did not expand on this despite the pronouncement that there is a reason. Maybe it was subtle, but then let it be subtle.

Have seen this so many times. And like you, having spent decades in tech, I know the cycle well. Some engineers know the problem today, but the directors will arrive at the same conclusion / concern 1-3 years later, when it becomes obvious that ... "oh this was wrong all along". But to be fair to them, they are dealing with a 1000 problems, not just this one, and that's where the management hierarchy you describe is completely liable (if someone could hold them liable).

I tried it once. Maybe 8-9 months back.I recall doing it soon after one travel, so that's why I can pinpoint it and this may be useful to figure which AI versions were available then.

The challenge is length and the context window. I had to own the plot since once it compacts, it has already lost the plot :)

I will give it pointers and tell where to go next. It will do it but either 1) stop after 3-4 paragraphs, or 2) write in a totally different direction than I expected. So I will nudge it to course correct, and it will do it. And then you keep iterating.

It was a good experiment for a couple of days. It saved me typing and it sometimes gave me lines of thinking that were interesting. But nothing that could be published.

But now that I think about it, I found a use case. If only GRRM could use its help :)

> no more skiing, endless smoke inhalation from wildfires, etc.

I wonder what our ancestors did, lets say 500 years back. Did they have wildfires? Skiing?

I get the point about humans causing unprecedented harm to the planet. However, the examples themselves are not perfect. I know skiing may be age old, but not as an activity enjoyed by millions, and the fact we build ski resorts may be contributing to some bad things, no?

There appears to have been a cease and desist issued by Corgi to Papermark: https://x.com/danielmerja/status/2070264877017350492?s=46

They want the tweet alleging copy to be deleted.

No I don't have a thesis.

About pricing, Netflix is constantly raising its prices and making more and more money (their stock is broken for last few months only) because they learned how to maximize profit by serving less customers but more profitable customers. Thats just one example.

All I am saying is that price stability is not guaranteed and is not a corporate goal so whether they failed here, whether they regret something, it is not possible to judge that just from price increase. Maybe they did, maybe they did much better than we know today and will find out next year (or quarter).

Begging? Component price goes up, they increase their prices. Where's the begging?

Just like US had oil but as the world prices went up, US oil prices went up. No one will sell it cheap inside the country for patriotism. Same thing here.

Of course they likely had / likely have some issues, but we should explain it as market forces rather than any struggle they are facing.

No me, well me too but not that bad, but my wife. Sometimes in the 200x era on green card, she will always get called for secondary inspection. Oh you entered this airport on this date. How are you re-entering without going out. "Well we did not. Not traveled for a year." But you did. All bags searched, more q&a and then they let her go. A couple of times they mentioned that the other person with the same name had the same date of birth. But the passport / green card number had to be different, no? I guess it took them that half hour to figure out and maybe the 200x AI matched by name and date of birth :)

But the sequel: a few years later I get a bill from a hospital for copay for delivery/childbirth. I call to contest ... we did not even live there any more, did at some point of time but years apart ... but they are adamant that my wife gave birth, at that hospital, on that date, in that city, and maybe she never informed me :) it was almost that weird. I don't act on it and give them a statement that it is not me/my family. Then another bill (or a final notice) a couple of years later. And then finally something clicks ... I used to work in a team where when I moved out, someone replaced me and his name was also same as me. Reach out to him, and his wife's name is same as mine, and they lived in the same city we lived in.

So someone somewhere fat fingered the wrong account when searching by name. He acknowledged the account (and childbirth) and paid up. I unfortunately did not ask him about his wife's date of birth to solve the immigration mystery.

My suspicion has been at my past employer's HR or legal department mixing up files

Let me see what I can do. Thanks ! May take me a month since I am sure I can polish it a lot more, not on how to create a deck but how to prep.

I am late 50s and I still find talks overwhelming.

What has worked though is

1. Recordings. Go into zoom, and give a talk with recording on. Listen to yourself, just once per recording but more if you can tolerate your voice :) It shows you your common failure points in black and white so that you can work on them.

2. Presenter notes. During my final talk, I have succeeded when speaking without notes, but to get there I need notes. I write my script as presenter notes (in Mac Keynote). Because of ADHD or whatever, as I am speaking, new ideas pop up. Every single time. In early rehearsals, I pause immediately, update the notes first and then go back and rehearse. After a dozen iterations, sometimes 30 iterations, the edits disappear, you learn to live with what you finalized already. Also once you narrate it 30 times, you have all kinds of memory maps sorted out in your brain, so even if you derail, you mind does have the information. I have seen some people have presenter notes in bullet forms, does not work for me. In the beginning I need my speech written out.

3. Rehearsals. The 30 rehearsals I mentioned. Figure out what you need. 5 or 50? You will know when you are ready. Once you have your magic number, stick to it. Maybe 25 becomes 20, but do not short change yourself by saying ... no I can wing it.

4. The morning of. I have tried 2 approaches. One is just open the deck and speak through the notes in your head. No recording. No zoom. No notes. Your mind should have it. OR not even deck. Just close your eyes, I am usually on a recliner (do not nod off) and run the narration in your head. Mistakes ... happen

5. Seating. Most of our presentations are in conference room settings but with remote viewers. Sometimes they are purely on video. Figure out the setup and rehearse 1-2 rounds in live mode. If you will be in a conference room, find one similar and rehearse once in that setting. If you will be seated and speaking to local + remote audience, try to visualize that as you present. It never feels the same with live audience, but it does give you some muscle memory.

6. Timing and Clickability. Transitions sometimes take more (or less clicks) than you envisioned. Practice them. If you are given 3 minutes, or 30, time yourself so that you know how far are you. Sometimes you need to be pitch perfect, but sometimes you can tolerate 10%. If I am at 10% variance, I stop fine tuning. As people ask questions, small variations can be tackled.

If someone else is sharing the screen and you are just speaking, rehearse that with them, because "next slide" handoff gets tiring very soon. To help them, add a line [ Click ] in the notes so that they know the click is coming when you get to a specific speech.

7. Screen setup. Over time, I am learning to speak from my mind (or heart). But the notes are there. I work on Mac Keynote, so there is a presentation mode where I can see what is on screen now, and what will be on screen next as soon as I click. The notes are there but they are only a last resort. If you are familiar with what is on screen, and what is coming next, you do not need to look at the conference room screen, OR any other window. These are usually enough to tell what you need. I make them the maximum size possible - do note that I have 2 screens side by side (current and next) so their sizes are still small on a laptop window, but having the biggest size I can get allows me to interpret fine variations - sometimes the next click makes only a small change, so having them visible helps.

8. Notifications off. Do not look at notifications. Ignore them. It is hard, but do it. 2 presentations back, one notification disrupted me so bad, I choked. I read the notification, it was at the beginning of session and someone just saying "sorry for joining late" and by the time I read it, my mind had disconnected. I started reading from the notes and never recovered from there. Maybe part of it was that I was not prepared - I did not have the 30 rehearsals. We did the work in 3 days, and 3 days was not enough for me to iterate on the slides, rehearse, have notes ... so when the chokepoint happened, my brain could not take it. But notifications is what triggered the disruption (of a bad kind).

Hopefully it helps

I think due to a variety of reasons EU and US are on a road to divorce. EU will be wise to make that happen on their terms, but they do not seem to be ready.

- Monopolies and related regulation. Of course US has its own companies being treated as monopolies so they will try to save them

- Social systems including healthcare

- Russia being next door vs far away (for US)

- The whole AI buildout

- A little bit of Libre Office smattering at the government level

Whether you consider US to be guarding its national interests, or whether you consider europeans to be taking a free ride on US's defense systems, it seems like 2 partners who got together for whatever reasons, you can even justify them in history, but history has moved on. I think we are going in 2 different directions for even the next president, however big a U turn he/she makes, I do not think the situation is salvageable.

Last comment and I will see myself out.

I meant my other productivity drops because I am not a natural presenter so even though I am rehearsing / editing for 2 hours a day, the presentation consumes me / overwhelms me that I can't even focus for the remaining 4 hours or 2 hours. Just do the bare minimum email processing, just survive. Everyone knows it. But by being in that zone of paralysis, I can still deliver a presentation. Sometimes good sometimes ok.

I have this need for the presentation content to reside in my memory cache and other work disrupts the cache quite badly.

But that's not a way to live. The other work stalled for 3 weeks.