> Just get off your ass and go and give them the message...
If I need to have all 4 members of the family meet me at the pool, first I need to go find each one of them. They could all be at different place. And then tell them individually to meet me at the pool? Is that the better solution you are proposing?
I work for a small company with a handful of devs. We don't have a dedicated devops person, so I do it all. Everything is self-hosted. Been that way for years. But, yeah, if I go on vacation and something foes screwy, the business is hosed. However, even if it were hosted on AWS or elsewhere, it would not be any better. If anything, it may be worse. Instead of a person being well versed in standards based tech, they'd have to be an AWS expert. Why would we want that?
I have recently started using terraform/tofu and ansible to automate nearly all of the devops operations. We are at a point where Claude Code can use these tools and our existing configs to make configuration changes, debug issues by reviewing logs etc. It is much faster at debugging an issue than I am and I know our stuff inside and out.
I am beginning to think that AI will soon force people to rethink their cloud hosting strategy.
You may not have meant to excuse the sad state we are in by presenting the "both sides are bad" argument. But it does have a strong whiff of it.
Both sides are bad. No doubt about it. It has always been that way. But, one side takes being bad to a whole new level.
Our choice has always been between bad and less bad. The voters decided to pull the lever for "massively bad" during the last presidential election because they could not tell the difference.
We get to give one bit of feedback to "the system" every four years. After four years of Trump, the feedback was "we don't want that". After four years of Biden, the feedback was "not that, either".
My impression of the US electorate is that they don't want illegal immigration, at least not in the volume and with the openness it was happening. They don't want immigrant trains rolling through Mexico. But they don't want the brutality and violence of the current crackdown, either.
They don't want trans people on womens' sports teams, and they don't want the US taking over Greenland.
And so on.
So after four years, the majority of voters were choosing "not Biden, and not the Biden things we don't like" rather than "yes Trump".
The place where it was "yes Trump" was the Republican primary. If you want to fix US politics, get involved with a political party - either one - and have some influence on who comes out of the primary process.
The politics of fear stoked by two sets of extremists egging eachother on is the core reason we're in this mess, the failure to reject both simultaneously and the desire to rule with feelings instead of facts caused it all.
I'm not a "whatabout" guy, I'm actively opposed to both extremes. The far left is just as capable of ruling with violence as the far right, they just haven't got the opportunity in this country yet.
The politics of emotion and absolutism is the cause, which flavor of extremism you pick isn't the core issue.
Except that the 'radical left' part of the electorate holds ~0 sway over the people who actually get elected. The 'radical right', on the other hand, has fully purged the GOP of anyone who isn't with their program through either primaries, or the fear of getting primaried.
The "radical left" stayed home in the presidential election, 9 million of them decided Harris wasn't radical enough for them so they would rather not vote, giving Trump the 2.3 million vote popular margin. Of course electoral analysis would be a lot more complex, I'm not doing that.
Elected democrats are stuck between trying to appease the radical left and trying to actually govern along with republicans and those two are very incompatible goals because the radical left knows very little about actual government policy and just has a couple of very narrow issues that most of the country opposes that their social bubble convinces them are the only important issues in this country.
Democrats lost the last election because of the radicals AND didn't get any of their goals done. Democrats needed a more centrist charismatic leader and instead they keep nominating candidates who "deserve" the nomination opposed to the actual will of the people OR continue nominating ancient relics who needed to retire a decade earlier... 6 of whom died in office over a period of 13 months in a period where every vote counted.
In short, I blame stupid leftist radicals and corrupt self-interested Democratic cowards in office for our current situation.
The "far left" statistically doesn't even exist in the USA. Less than 1% of the population and less than 0.01% of elected politicians. Effectively zero. No major national or state politicians call for seizing the means of production, a centrally planned economy, widespread price controls, Great Leaps Forward, and so on. The far right has us all convinced that anything to the left of Reagan is "far left".
You can pretend all you want. The proof is in the reality of what we see. Mass illegal immigration is not a center-right or even left position. It's a "far left" position that has had enough political power to be enforced for years now.
That's the most disingenuous argument yet, that completely ignores reality.
The libertarian party does NOT advocate for illegal immigration today. Only once NAP-violating governments are abolished, do they contend that the harmful effects of unconstrained immigration are non-existent. They argue that if you want complete freedom of movement, you must first give up all claim to nonconsensual government force, including taxation and redistribution of wealth. And even the most radical of libertarians include a "non-aggression" exception. They support screening to exclude violent criminals, security threats, and health risks. All of which are impossible if there is no monitoring or reasonable enforcement.
You know _damn well_ that in the context of actual politics in America as they stand TODAY, open immigration is a FAR LEFT position. It's undeniable to anyone arguing in good faith.
The Libertarian Party does not put all of those caveats on their stance and you sure as hell didn't in your polemic post.
Your argument is nothing but pedantry anyway. The libertarian stance is that most immigration laws should be abolished. Whether or not the books on the laws reflect their stance doesn't change that is their stance. They advocate for open boarders and migration free from government interference.
>The politics of fear stoked by two sets of extremists egging eachother on is the core reason we're in this mess, the failure to reject both simultaneously and the desire to rule with feelings instead of facts caused it all.
Pol Pot[0] was a leftist extremist. Chairman Mao[1] was a leftist extremist. As were the Red Brigades[2] and the Symbionese Liberation Army[3], etc., etc., ,etc. Who in the US Democratic Party advocates for the same things as those guys? Let's see. No one.
In fact, the only ones in the US who've shown an interest in nationalizing the means of production (c.f. Intel) or putting down the Intelligentsia and normalizing violence against those who criticize the regime are just one set of extremists. Because extremists end up going full circle -- because for them it's about power and not ideology.
Computer use and screenshots are context intensive. Text is not. The more context you give to an LLM, the dumber it gets. Some people think at 40% context utilization, the LLM starts to get into the dumb zone. That is where the limitations are as of today. This is why CLI based tools like Claude Code are so good. And any attempt at computer use has fallen by the wayside.
There are some potential solutions to this problem that come to mind. Use subagents to isolate the interesting bits about a screenshot and only feed that to the main agent with a summary. This will all still have a significantly higher token usage compared to a text based interface, but something like this could potentially keep the LLM out of the dumb zone a little longer.
More to the point though, you should be using Agents in Claude Code to limit context pollution. Agents run with their own context, and then only return salient details. Eg, I have an Agent to run "make" and return the return status and just the first error message if there is one. This means the hundreds/thousands of lines of compilation don't pollute the main Claude Code context, letting me get more builds in before I run out of context there.
I was at a Bills game in Buffalo last year. 5 rows ahead of me was big tall dude, who stood up and would not sit down. This was blocking the view of everybody behind him. People grumbled, but nobody said or did anything for about 20 minutes. I was quite peeved. Then an old lady right behind him gently tapped him on his shoulder and reminded him that he was blocking the view of several people behind him. The dude shrugged his shoulder and said, "not my problem, you can stand up too if you want". I am a mega-nerd, but I lost my cool right there and then and started screaming at the guy. My girlfriend, who didn't want to see me get beat up, pulled me away from the scene.
Many people are just massive assholes. Asking nicely does not work. Particularly big drunk dudes at an American football game. That was my first and last visit to a football stadium.
That really sucks, but don't deprive yourself of something you think you might enjoy because of that one jackass. Chances are that next time you won't experience something like that.
Also, basically every pro and semi-pro sports stadium nowadays has cell-phone-contactable security that you can summon to handle situations like these. The threat of being kicked out of his $250 seats is way more of a threat than that of being confronted by a "mega-nerd".
I wouldn't make a habit of contacting security over every little annoyance, but if they're obnoxiously blocking an old lady, that's the time to use it.
This is good work. When a task is of critical importance, I give two different LLMs the same task. And then ask them to review each other's output and validate all claims. I do this with Codex and Claude Code. It is very rare for them to find some valid fault in the other LLM's solution. And they are generally good about admitting mistakes and then creating a single unified solution that addresses identified issues. This result is better and ready for human review.
I don't have a college degree either. I am about 50. I have never been unemployed and have had high paying software dev jobs my entire adult life. Your claim that the lack of degree is the only thing holding you back is very much incorrect.
I suspect the problem is elsewhere and you are unwilling or uncomfortable to discuss it.
I concur. I tried oh my zsh and for a while I thought I was doing something wrong. My experience was terrible. But it turns out that many people did not mind the bloat and the prompt lag. It was not for me. I uninstalled it and I have such a bad taste in my mouth that I am very reluctant to try any of the alternatives.
I have been using a 48" LG OLED TV as a monitor for about 2 years. I thought I would love it. But I hated it. Text looked horrible. I was going mad and then Google'd a bit to see if others hated it too. And found that they did. But luckily there are settings that can be changed to turn it into an excellent computer monitor. Once I changed the right settings, it was love. I have 3 monitors on my desk. 32" LG LEDs on the side, 48" OLED in the middle. All 4K. I love this setup. I do occasionally think about replacing the LEDs. I just need the OLED pricing to drop a little more.
I did use whisper last night to get the captions out of a video file. The standard whisper tool from OpenAI uses CPU. It took more than 20 minutes to fully process a video file that was a little more than an hour long. During that time my 20-Core CPU was pegged at 100% utilization and the fan got very loud. I then downloaded an Intel version that used the NPU. CPUs stayed close to 0% and fans remained quiet. Total task was completed in about 6 minutes.
NPUs can be useful for some cases. The AI PC crap is ill thought out however.
Depending on the part, it's likely the iGPU will be even faster. The new panther lake has iGPUs with either 80% or 250% the performance of the NPU when at the higher end. But on lower end models, it's lower but still within the same performance class
Batching is essentially running multiple instances at once, ie bundling 8 segments and running them simultaneously on the processing unit, but which obviously takes more RAM to do. Notice, however, that if you drop the precision to int8 from fp16, you use basically the same amount of RAM as whisper.cpp yet it completes in a fraction of the time using batching [0].
Yes, if you check their community integrations section on faster-whisper [1], you can see a lot of different CLIs, GUIs, and libraries. I recommend WhisperX [2], it's the most complete CLI so far and has features like diarization which whisper.cpp does not have in a production-ready capacity.
If you mean OpenVINO, it uses CPU+GPU+NPU - not just the NPU. On something like a 265K the NPU would only be providing 13 of the 36 total TOPS. Overall, I wish they would just put a few more general compute units in the GPU and have 30 TOPS or something but more overall performance in general.
If I need to have all 4 members of the family meet me at the pool, first I need to go find each one of them. They could all be at different place. And then tell them individually to meet me at the pool? Is that the better solution you are proposing?
reply