phil21

While I certainly don't work at Cisco, I've come to the conclusion that I need to re-examine my priors when it comes to code rewrites and general maintainability.

AI makes these projects significantly more feasible, at least in my domain. We have some critical-path software I wouldn't touch with such an approach, but the vast majority of our code is internal and/or supporting tooling.

For those cases, burning it down and rewriting via AI tooling has proven far more productive than anything we have attempted in the past. This is for old technical debt laden stuff everyone has been afraid to touch for well over a decade.

It's not perfect, but it moves the needle where before a lot of such projects were stuck on the backburner for years.

It certainly introduces new bugs and such, but at least this way forward momentum is happening vs. just being stuck in decision/analysis paralysis forever.

This also assumes some very senior SME folks are driving the process. I would not expect similar results with junior or even mid-level folks in these specific domains.

> Weirdly, underutilization of hardware was never a problem before.

Fixing exactly this has been the constant drumbeat of IT since the early 00's at least. It started with virtualization, to get capacity factor of hardware up in the enterprise. The average server back before VMWare (and HyperV/etc.) became a major thing was probably sitting at 10% utilization or less 95% of the time. Hypervisors were invented to fix this as their primary use case. You'd take a 5-10% performance hit on the raw server performance to gain 90% efficiency back.

Next came the cloud. Which was virtualization at scale. The larger the scale, the better the efficiency.

It's basically almost all anyone has actually accomplished in IT over the past decade or two. Everything else has been largely stagnant compared to the 80's and 90's.

AI inference is in it's infancy, but I imagine we will see more or less the exact same thing play out yet again. Lots of initial on-prem stuff due to cost and freedom, and eventually it becomes virtualized, commoditized, customer segregated, etc. Just like average cloud workloads are today and all the various "small" megawatt-scale builds moved back into the hyperscaler type providers within the next 5-10 years.

> How many companies are building datacenter commons and sharing with other companies to optimize resource usage?

Quite a few that I've seen. Rented inference/token factories are the current new hotness. Plenty of speculative builds being done by hundreds of smalltime providers. These will consolidate per usual and everyone but the last few top percent will remain as the winners. Just like the first hosting providers. Virtualization and fancy schedulers to maximize hardware utilization across a number of end-user customers are a huge topic in this space at the moment. Heck, a lot of these small time players are providing raw compute for the giant AI labs and even hyperscalers.

I don't like that Altman is right on this btw. I much preferred the highly inefficient days of every midsize corporate office having a 200-500kw datacenter in the basement. AWS and their ilk are quite obviously orders of magnitude more hardware efficient though.

No, it's recognizing mistakes will be made and no one is perfect. You will also make mistakes regarding blast radius - sometimes the blast doors don't close or function like you thought they would during an actual explosion.

Defense in depth is the idea. Swiss cheese model of failure.

Very much agree with most of this.

The number of people who never seem to have understood what software and IT is in most companies is eye opening to me. Your job was never really to "write code" - you were always automating things.

That the field got somewhat stuck for a couple decades seems to have made many people forget this.

The vast majority of programmers who started in the 80's and 90's are basically directly responsible for massive waves of job losses in certain categories. Lots of cases of entire departments of secretarial level paper-pushing work being replaced with a small team of programmers. Same will happen with AI tooling. The difference is programmers are no longer the "winning" party here.

Computers and software in the end are the most configurable and complex automation system society has ever created. They are simply a continuation of industrial automation, and AI is a continuation upon that.

Been a long long while but basically empty flights were semi-regular things back in the day if you booked what you knew to be repositioning flights at odd hours.

I’ve been one of a handful of passengers more than a few times. Where the flight attendants just move everyone to first class after boarding and the rest of the flight is super relaxed and barely professional.

This was not common, but not a rare thing after the 2008 crash for a few years.

During IRROPs you can still luck into one of these. Definitely a fun experience!

> Few people in a neighborhood of single family homes would want a 20-story apartment tower getting built next door.

Only this is just a thing that really doesn't happen much at all. Very few developers are putting up 20 story apartment buildings in the middle of an existing SFH community in the suburbs somewhere. Most of those such fights are people fighting the 5 over 1 development built greenfield just outside the current SFH neighborhood in question. And for those such people? They should have zero say.

If they are doing so, it's typically in very dense city cores that support the density and thus investment. If someone wants to put up a 20 story building next to me in my Chicago neighborhood, sounds good to me! I am upset there are not more of them and that zoning has effectively made them illegal to build here, as I'm roughly 500 feet away from a rapid transit station and two arterial roads with high frequency bus routes. It's a crime that they are illegal to build.

I'd be an asshole if I bought a house in such an area and wanted my little conclave to be protected at the expense of everyone else in the city. I should rightly be overruled and told to pound sand and move the hell out if I don't like it.

Instead, laws have enabled such people to effectively veto such buildings. Or even simply stuff like 12 unit 4 floor apartment buildings that already exist due to predating zoning on my block.

Many areas of the US are effectively retirement communities by zoning. People who moved in 20-30 years ago who demand zero change and have captured nearly all political power.

Not my experience at all in my neighborhood. Development gets torpedo'ed by current residents. Parking and traffic almost always being the primary one. As in, more parking minimums. Developers showing up are the only people trying to actually build anything at all, most community members are NIMBY.

Not sure what a private developer would have to do with public transit though. That's the city's job to provide and figure out the routes for. Developers bring projects accordingly, and the local residents veto 12 story apartment buildings built on current strip-malls 2 blocks away from train stations. Residents using transparently bullshit environmental and walking/biking infrastructure "concerns" as just another way to block things.

I'll take building stuff - even if flawed - over the current state of never building anything ever any day of the week.

> and be spared of the implementation effort

I enjoy the effort - most of the time. I don't enjoy the time burned on what amounts to a very small portion of the entire chain of creation.

Shrug, assume bad faith if you want to. It’s exactly how I read it when it was first marketed. I didn’t think it met that definition then at the time, but these days it’s pretty close. Plug in a destination and it works until it hits an exception it can’t handle.

Autonomous driving is something entirely different in my mind. That would be a car good enough to work without driver controls or a remote operator. Nothing on the market meets this definition today.

In that case, autopilot is the correct term.

Pilots still are required to be able to take over immediately for an autopilot system and are responsible and accountable for all those items you list and more.

Calling it autopilot is precisely what it is, calling it self driving would be false marketing.

I would usually have been part of the downvoting brigade for this comment.

But to be honest the reaction from the ivory tower here is seemingly quite telling to a layman watching from afar.

It looks far more like this was some extremely gatekept community where you were only allowed to work on sanctioned things if so deemed worthy. Better not step one someone’s toes lest you get exiled! How much progress was lost due to newcomers never working on a problem due to told not to?

It’s starting to make those old timey Usenet kooks look a whole lot less crazy to me. They seemingly were right about this part.

Mathematics was funded and supported by a whole bunch of janitors and blue collar workers busting their asses to let entire departments of thinking men and women work on hard problems for the betterment of society. If it was just a super special friends club the whole time that was gatekeeping what amounts to a hobby, it’s going to me much harder to justify such a social contract.

Honestly not even sure what to think at this point. Extremely disappointing to me the more I read from these people.

Sure, but their energy budget includes everything. They don’t have the luxury of importing energy others invested into decades ago when convenient and having it be off balance sheet like Denmark does.

China would actually have to build stuff to do that. Denmark can just pretend it doesn’t exist.

Not how it worked in my home state growing up. The state intermittently ran budget surpluses and were net givers to the federal budget. Still are today.

I’m also not entirely sure how relevant that is due to states (and provinces) not printing their own currency though. Very different situations.

You don’t make meaningful progress when you focus on a rounding error part of the budget. It’s like targeting private jet use for climate change. It’s for the feels so you can pretend you’re on the right side of things while doing absolutely nothing that matters in final effect even if 100% successful.

If you want to solve the budget deficit you need to go after the things that matter which is mandatory social spending these days. That and the revenue side combined are all that matter. Even the military is starting to become more and more irrelevant as social spending and debt service eats up all the air in the room. You could end it entirely tomorrow and you’d still have a deficit. It’s not the 1990s anymore.

Anything else can be immediately written off as unserious action brought by unserious people. The bike shedding of budgetary politics.

Depends where you fly out of mostly. When I lived in MSP Delta was both the only real choice, and also pretty damn good for an airline. Especially with top tier status.

I moved to Chicago and am stuck with United now. I miss Delta most times I fly, they simply had a better product.

Same. My theory is the ginger in them. But who knows!

They wouldn’t mine with them, they would just offer cheap inference with open models.

This is already happening.

Good luck powering a 400kw rack in your basement, if you are looking to do frontier model type stuff. Just a single machine can run in the 150kw range these days for the less exotic stuff.

It can certainly be parted out with proper motivation but it’s going to take a whole new ecosystem to support these in your typical home lab. Even then 20-30kw slices are electric car level energy consumers.

> do they plan to increase production? No.

Micron has 3 brand new fabs currently under construction, 2 Boise, 1 in New York as the first of 4 planned for a campus.

Plus expanding other existing facilities.

These things take ~3-5 years from breaking ground to full production. You'd have had to anticipate the current demand years before it happened in order to be bringing production on-line before 2030 or so.

Samsung and HK Hynix also have fabs under construction and planned.

CXMT started 11 years ago and only now is reaching any real volume. If they decided a year ago to react to the current demand cycle they'd be 6-7 years out.

Not much you can really do to wish for more fabrication to exist on any timeline not measured in fractional decades.

Could they do more and react quicker? Probably, but everything I've read on the subject seems to point to 3 years is absolute bare minimum if you happen to have a shovel ready project with the land bought, local permitting completed, infrastructure extended to the site, and a skilled workforce already in place. They could suspend buy-backs/dividends today and dump it all into building production and there would be no material impact until around 2030.

> The Micron CEO just recently said this is the exact plan

CEO simply stated the demand pressure will not go away through 2027, and supply will not increase until around 2028 when currently under construction fabs start shipping volume. The article does not support your statement.

I had a list of controllers that we forgot the usernames and passwords to. We knew it was some combination from about a half dozen of each, just not exactly what was what. A few hundred of them so a tedious task for a human to go through and validate them all.

Got flagged for attempting to have Codex write a quick script to basically dictionary attack my own infrastructure with a tiny dictionary file.

A bit silly, but it doesn’t take a whole lot.

> So your inability to supply that demand resulted in you creating the conditions for competition to show up, and that competition can very likely outcompete you out of the market.

Doubtful. Your competition started building out additional production roughly when you did. Lets say you got a 1 year head start on shovels in the ground.

Those competitors have decades of experience producing RAM. You do not. Even if you get your facility on-line to first wafer a year earlier than the incumbents, it's unlikely you will beat them to full production output as even extremely experienced operators take 6-18mo to get a fab from first production to expected nominal output.

The only way to "win" this is to have started before the supply crunch happened back in 2021 or 2022 - or more likely much earlier. Like CXMT did.

If you were able to find a bunch of investors to give you $40-50B today, already had a site ready to start construction literally tomorrow, by the time you are on-line you will be competing with new capacity currently being built by Samsung, Micron, and SK Hynix - not to mention CXMT. The only way you will be profitable is if the current demand exists through 2031 and beyond.

I know I certainly wouldn't be investing in such a project, since the entire bet is that RAM prices 5 years from now will still be at roughly what they are today. Which is simply a bet that the current AI bubble and infrastructure build-out is still ongoing.

> But here's a CEO saying he's going to make sure prices stay high enough, long enough that you can spin up a competitor

2 years is not remotely enough time to spin up a competitor. It's not even enough time for a current leading competitor to spin up another fab.

If you are a newcomer you are talking a decade or more before mass production. If you are HK Hynix perhaps 5 years or so.

The majors all have large fabs under construction, but the majority of new production from those facilities won't be coming fully on-line until 2028 at the earliest. Even from the date of first wafer shipped most of these facilities take 6-12mo to reach mass production level output.

Chinese DRAM is interesting only because they started building capacity (expertise) a decade ago. Only now are the expected to have any material impact on the total volume of the market. And if they want to gain even more market share, they will be under the same ~5 years to spin up a brand new fabrication plant. Perhaps a bit less because they are China, but still multiple years.

It's a loose RSVP. A few extra folks showing up last minute is not a big deal, the meals are usually "family style" and leftovers get taken home most of the time.

If someone is rolling in with a larger than usual group they typically bring enough extra food to go around.

And yep, a lot of these are pot luck style to some degree. Host makes main course, everyone else pitches in on other stuff.

Definitely some loose coordination involved and the host will typically call a roll call the day before to get an estimate of number of people dropping by.

And it's not just a few people. The sophisticated scalping operations have hundreds of straw buyers in their network. For items that generate far less profit than a 5090 currently does.

I've watched vans dropping of an endless stream of people in the parking lot of my local Costco to buy out all the cases of Pokemon cards within 20 minutes of the "random" drop they put on the floor. They go in, buy the max quantity of the item with cash provided by their keeper, bring it back to the van with a receipt and change, and get the $20-30 per item pay they were promised. There were 2 or 3 'legit' buyers out of hundreds of obvious straw buyers.

If the profit per item is $4-5k, expect this to go into turbo mode and be done in a marketplace/automated fashion and go viral. Just sign up to your favorite scalper's platform, submit your information, get a quote and a prepaid label and drop the box you got from nVidia off at USPS the next day. That "tens of thousands" of individuals will turn into "millions" really quick once it gets known there is rent money involved by clicking a few buttons and it goes viral on all the social media platforms.

The only way out from that is manufacturer enough cards to meet demand, and at that point the retail channel doesn't really matter.

I helped implement the concept of a "Sunday Supper" with our friend group in the city. It's basically held every Sunday and rotates to different households who feel like hosting that day. Luckily one person is sort of the "default" since their family's child situation makes it more convenient for them to host vs. go elsewhere.

It's basically an open invite for a group of friends, and a standing thing that just happens. If you have no other plans, you basically go. Sometimes I have to drag myself out of the house to do it - but it's worth it every time. The group that attends sometimes is just one other couple, sometimes it's a house full of 30 people if everyone comes, brings their kids, and a few friends of friends tag along.

Definitely one of the better things I've been involved in!

> Especially when they cross into the physical realm as in not properly secured and air gapped control systems. SCADA is scary.

What about the bad actors (choose your own evildoer here) who purposefully do not air gap their agents? And specifically train them to attack in such a manner?

I'd much rather have relatively benign stuff like this hit first, because the former is coming sooner than later. It's already here in a limited manner, likely more than any of us currently realize.

Botnets could crack passwords faster than anyone thought possible over 20 years ago now. This is just the latest iteration of such a concept.

There is so much low hanging fruit in this space that frontier models are currently utterly irrelevant. It's going to take decades of human-speed securing of IT to make superintelligence or whatever you want to call it a necessary component for such attacks.

At this point, someone with a rack or three of GPUs with 100kw to burn can replicate such attacks if they feel like it. the bar for entry is not even 7 figures.

It’s almost always both. Basically every single retired person I know in my family or friends of family are relying on their home to be their primary asset/income in retirement. It’s always a much larger chunk of their net worth than any retirement savings, if they even have the latter.

They were commonly called forced savings accounts when I was growing up. The standard advice was to buy a home as an investment vehicle. Every older family member giving me advice would preach about how they don’t make any more land and the price will only go up.

You can - to this day - talk until your are blue in the face to these people that your primary place of living is not an investment, but they will refuse to even entertain the idea even with the stark cold math staring at them in the face.

If these retired folks can’t sell their homes for a huge profit they will be in for some very difficult times a decade or two from now. They are highly motivated to keep the music playing as long as they can vote and show up to local city council meetings.

Rehabilitation has never really been the point of the US prison or even correctional system at all.

The fact you can serve your time, earnestly work on rehabilitating yourself, jump through all the hoops, etc. but still be under the economic death sentence more or less summarizes how people feel about it.

It’s basically NIMBY applied to the justice system. Everyone loves to pretend they want rehabilitation but also wants someone else to do it. This shows up in hiring, apartment leasing, social benefits, etc.

> The U.S. locks up people at 5x the rate of South Korea. Whatever the difference is, it’s not because the U.S. is unwilling to punish people.

The difference is in consistency in enforcement and general background crime rate.

If you have a nearly guaranteed chance of getting caught doing something, the punishment almost doesn't matter. If every single time you went over the speed limit you got a $500 fine 100% guaranteed, there would be effectively zero speeders ever. If you have a 1 in a million chance of getting caught but the punishment is 10 years in prison - there will still be plenty of speeders all over the place.

Humans do not do well with estimating risk on low probability events. We are all just big children in the end. Consistency in consequences are what actually matters, not overly harsh punishment.

There was/is a loophole where if you signed up via the iOS or Android app, it allowed it.

It's been disabled for some time now though otherwise, I check about once a day myself and keep and eye out on social media.

Annoying since I was about to upgrade back to the $200 plan after downgrading to the $100 plan due to being on leave and not needing as much usage the month prior. Doh.