"We spent $10,000 in tokens last month, but are we sure we're seeing a proportional increase in business value?"

It's a very good question and one that I think is going to decide whether agentic ai for coding remains a viable use case, it's possible that it's just too expensive to be worth it.

OpenAI and Anthropic want you to think of their agentic products as virtual employees, so when you think of it that way $10k a month becomes a single salaried employee at $120k a year, with no healthcare or other administrative costs it's practically a steal.

But on the front lines, actually working with LLMs day in and day out is like working with Amelia Bedelia (iykyk) if she were a brilliant computer science intern instead of a housekeeper. Nodding and smiling away as you feed it detailed prompts and responding convincingly as if it understands, and then goes off and thinks of a very convincing plan of action, only to continually miss the forest for the trees.

This is okay I guess if your employer is blindly footing the bill at one of the larger companies, but I don't think it's sustainable long term for companies realistic limits on their finances.

In my personal projects it's definitely caused me to change how I use agents, when Github Copilot was first released I was using it all the time and barely able to crack 50% of my monthly usage limit. After the recent changes where multipliers for usage were applied to frontier models I was barely able to use it for a few days before hitting my limit.

Their new pricing model feels even worse, $40 a month where they give you $40 worth of tokens a month doesn't seem worth it, I'm pretty sure I can blow through that in a day or two of normal use. The worst part about it is that I don't feel like I really have control over how that money is spent, if I ask an agent to do something and it spends a million tokens completely missing the point, should I have to pay for that?

Taking a step back the whole thing starts to look suspiciously like a slot machine. I pull the lever by entering a prompt, and the dials spin for a bit before producing code output that's almost a jackpot, so I refine my prompt and pull the handle again, and again, until the code does what I wanted it to. Probably a gross oversimplification since the output is not completely random, but randomness that is guided by model weights is still random.

OpenAI just demoed their latest model SoL (shit outta luck wouldn't have been my first choice) and two things stuck out to me.

First, they demoed a paper plane game where you guide a paper plane through rings as it glides down a boring hallway Pilot Wings 64 style. Right off the bat my first impressions was that it was very janky looking, graphics were clipping, objects were popping in and out of the environment. Just a poor game overall, for such a simple concept I would have expected their best model ever to be able to at least give me a clean demo.

So the next thing they demoed was using an agent with vision to fix some of the bugs in the game. I'll point out that they didn't choose one of the obvious visual glitches to fix, but they changed some text on the screen to show that the model could do things like update the text in a dialog. Okay, that gets a golf clap from me but uh, why was this simple demo so buggy in the first place?

Second, they showed off a demo where they wrote down some requirements and fed them into the new ChatGPT app, and then watched it spin up agents to write the tickets and then more agents to work on the tickets, but they didn't show us the final output which seems like...a glaring omission no? Maybe I'm just being cynical, but I feel like the folks who have the most access to the best models should be producing tighter demos.

I think the writing is clearly on the wall, I'm certainly not going to pay $100s per month for a robot to write code badly. If most people feel the way I do, it means that we're most likely going to see a shift to local models for most development unless you're working for a company that can afford to pay. I'm pretty sure that's not sustainable either though, unless the quality of the models gets dramatically better very soon I think you're going to see a falloff in adoption just due to sticker shock, and no one living off of this bubble can afford that.

The bones of Zero Company are immediately apparent. It certainly looks just like one of the latest entries in the XCOM series with a Star Wars paint job, and if that's all it was it would probably still be a pretty decent game. There is more too it though, the team over at Bit Reactor appear to be fans of Marvel's Midnight Suns and the way that game introduced a hub world, an X-Men style mansion, where the heroes could interact between missions. They don't lean into it as much, but Zero Company's HQ back at the Ring of Kafrene does the exact same job of pulling you into the world and letting you see how the team spends their off time.

Similarly, in longer missions where there may be two ore more battles strung together, the player gets to guide the squad between the encounters. Now, these are just glorified hallways to be sure, but the alternative would be a simple instant transition where your squad runs off of one screen and into the next battle. Being able to skulk through the beautifully rendered back alleys and space docks does a lot to pull players into the game, and they didn't have to do it so I appreciate it.

You can see this level of polish throughout the game, after you've created your main character, Tesh Hawks, using a decent but not extensive customizer you'll find that when the game loads up to one of a few random idle animations of you chilling in your office. Usually the kind of thing reserved for a loading screen, so I often find myself waiting for the game to load in before remembering that it's actually waiting for me to make Hawks stop scrolling through his financial reports and get to work.

The voice acting is all top tier as well, I genuinely look forward to walking through the base after each mission to touch base with everyone on the squad. The game includes a handful of named squaddies, or operators as they are referred to in this universe, that have interesting backstories and side quests that you can complete. But I was surprised to find that even the operators that I customized and hired to fill out the team have interesting, albeit slightly more generic, things to say as the story continues to progress.

Another element that Bit Reactor has added is called Bond which represents team cohesion, the more operators team up the higher their bond, and having a high bond will unlock "cross training" opportunities where you can select a single stat boost that will affect both bonding characters. Effectively this is one of the two ways that characters level up in this game, the other is by using focus points where you distribute points through a standard tech tree, so it's easy to sleep on Bond and not realize it's there.

The last feature I want to talk about is the Call for Backup ability. Each round a random operator on your team will get this ability and they can use it to target another friendly operator so that when the original operator takes a shot, the second one will follow up. Pro Tip: The second shot is always guaranteed to hit as long as the second operator has any line of sight on the target. This is wildly powerful when used correctly. On top of that, every time operators back each other up it boosts their Bond, even if they don't end up firing a shot.

For Star Wars Zero Company polish is the name of the game. Bit Reactor could have easily slapped a coat of Star Wars paint on a bunch of menus and called it a day. In fact, if you hate walking around the base and talking to people you can skip all of that and open up the base map to directly access anything you need without ever leaving Hawks' office. When you look at it that way the entire hub zone is just a flex 💪🏽

Now for the bad news, because I can't in good conscience end this glowing review and not mention the fact that nearly the entire team that worked on this amazing game got furloughed a few weeks before it came out. You would think that when you're making a game like this, working with Disney IP, that you'd have some kind of job security, but no one is safe in these strange days. It can't be good for the industry, I grew up wanting to make video games for a living but I saw the writing on the wall 20 years ago when EA started the trend of releasing a game and then laid off a chunk of Tiburon Studios, their main development house, and it seems things have gotten even worse.

So on that note, go buy this game and hope those folks get rehired to make some DLC, and have a great labor day weekend.

I'm at peace with that, I'm fortunate that although we don't see eye to eye on many topics, we've always had an open line of communication for as long as I can remember and she has always been one of my favorite people to talk to. These days each conversation has the feel of a post dinner chat at a fancy restaurant where the staff is much too polite to kick you out but the cheque and tip has long since been picked up. You suddenly remember all of these things that desperately need to be said before parting. We stretch these moments out as long as we can, long strands eventually becoming wisp thin and breaking with a soft pop, with luck etched into our memories forever.

Yesterday we salvaged the day as best we could, I watched Rush Hour with Brendan and Oli, took a long walk with Rubi, and later we capped off the evening with a family game night of Jackbox 3 and 7.

Today is Shardé's birthday and I'm making her breakfast in bed, no really big plans but the boys are all excited about the gifts they picked out for her so we're all looking forward to that. I have to hop off to get started, 'til next time.

I've seen a few instances recently where Twitter or YouTube users that I know aren't bots per se, are posting tweets or reading scripts that are obviously AI generated and then lying about it. I've been thinking about it and I've come to the conclusion that the only way a reasonable person would find that kind of lie defensible is if they truly can't see what is obvious to everyone else.

I'm sure there are people doing studies on this already so I should go look it up, but I wonder what the percentage of folks is right now who can't tell in a paragraph or two that a human didn't write something? I don't mean people who can't tell most of the time, I mean the people who can read an LLM generated book and upon completion hand it back saying, "It doesn't look like anything to me", West World style.

I'm probably making a bigger deal out of something that is likely to be a transient problem, someone will eventually figure out how to make LLMs sound more generic.

But what if we can't figure out how to cross that uncanny valley? Will we end up with a subset of people who all use "AI voice", essentially meat proxies for ChatGpt?

Something to think about, this probably should have been a Tweet but Elon ruined Twitter so you get this, cheers.

Now that I'm back a quick glance at the headlines shows that I haven't missed much at all, the war that I predicted would become a forever war all the way back in June continues to grind on. Today's headlines are about how the US Navy is breaking records for time at sea, sailors jumping 60ft off the sides of their ships attempting to self harm.

There is a cyclospora outbreak sweeping the country, it's been generally linked to lettuce but no one has gotten to the source, other basic food items are being recalled, meat packaging facilities are shutting down because there aren't enough cows. I'm not cherry picking headlines at all, it seems no matter where you look the news is bad and there is no light at the end of the tunnel.

I'm past all of this stuff enraging me, rage has it's place but being constantly enraged and doing nothing just leads to burnout. I have to dial back the rage and settle into a deep dissatisfaction while also pushing back against the way that Trumps constant firehose of lawbreaking is becoming normalized and calcified in the presidency.

Trump and his team do not move like people who are worried about being prosecuted after they lose power and I think it is because they never expect to lose power. As this continues to play out what I feel the most is a sort of sadness mixed with fear, sad for what we've already lost and also afraid that they might be right.

They've also both done pretty well in swimming lessons, I'm finally less worried about losing Bren to the waves at the beach.

Oliver spent a week at sailing camp and I'm really proud that he survived being capsized and was able to right his little boat and get himself back in it, great stuff all around.

Negative points for not getting out to the pool and stuff as much with Miles so I'll have to pay extra attention to spend the more time with him at the beach this week, he loves the water as much, if not more than I do.

He's been making pretty great strides behavior and communication wise and while that's great it has a way of letting him blend into his routine in the background which is less great. Which reminds me, he's been doing really well with this cheap off the shelf word board that Shardé picked up for him, it's very simple and only has a few dozen words but since there are no apps to distract him the battery is never dead so he's always able to use it.

I think that we can make something very similar with a piZero or nano or something that could compete with the ipad apps but with the added benefit of having no other distractions on the device, best of both worlds. Something to work on when we get back from the beach.

Getting back to our busy summer though I want to make sure to mention our trip to the Nats game. Derick from KLA invited Oli, Brendan and I to come see the Yankees at the Nationals Park in DC so of course I said yes and dragged the kids somewhat reluctantly along for the ride.

The trip there was mostly uneventful except for it being Oli's very first time on an extremely packed train, probably terrifying for him and if I had to do it over again I would wait for a car with fewer riders. In the moment I pushed us all onto that train despite it being over capacity because I was worried about being late to meet with Derick and there was a rush of people behind me also trying to get on, but no excuses. That one is on me.

I would like to try to go again and either go earlier, wait for a less packed car, or both.

Those are the highlights that I can remember off the top of my head at, we'll call this the halfway point of summer, pushing it I know, but the MLB all-star break was just a couple of weeks ago so I hope that the judges will allow it.

Going into the second half my goal is to keep the momentum up, get the kids back on their bikes, spend more time with Miles, work on some simple robotics projects, take lots more photos than I have been, and try to make as many memories as I can.

I'm going to truly try to disconnect for a couple weeks, in the past I've only unplugged from work while continuing to keep up with the news, but I need break from the daily outrage machine.

Signing off for now, stay safe out there.

Before reading about their experience I was a firm, almost dogmatic, skeptic about anything of the sort. People who believed in UFOs were in the same camp as people who hunt ghost, bigfoot, and the loch ness monster, or so I thought.

After looking into the subject, something which I strongly invite every skeptic to do, I found that there are a few key differences between UFOs and other unbelievable supernatural sightings. First and foremost being the large number of sightings by credible witnesses. Second would be the extremely suspicious behavior of the US government around the topic. We've spent millions of taxpayer dollars over the decades investigating the topic, while at the same time saying that there is nothing worth investigating.

You can't say the same thing about ghost, or bigfoot.

Over the last few years there have also been a handful of congressional hearings on the subject of UAP, you would think that if there was nothing there that this would quickly be cleared up with a top secret closed door meeting or two, but in fact the opposite has happened. There have been several top secret meetings in secure facilities and while the congresspeople can't say what they've seen it is very clear that they are coming out with even more questions.

In recent months the increasing political pressure has gendered a response and the Whitehouse has directed the DoD (Department of Defense) to begin declassifying and releasing UFO information. The announcement was made in kind of a tongue in cheek way, and because Republicans are generally seen as anti-science and the kind of people who believe in bigfoot and ghosts anyway it still hasn't gotten the attention of the general public.

As of yesterday the government has dumped four tranches of content, one every month or so since the announcement. There are lots of documents and several dozen videos, mostly low quality military grade stuff that you would need to be a trained observer to interpret. There's definitely some weird looking stuff in there, but nothing that I've seen so far that feels definitive.

A couple of weeks ago the notable skeptic and very popular public science figure, Neil deGrasse Tyson quite suddenly changed his tune from, "If UFOs were real, why in the age of the iPhone don't we have more than a few blurry pictures of them" to, “The real question is not ‘are we alone?’ it’s ‘are we ready?’ Are we?”

When asked why he is suddenly more open to the topic he cites a recent shift in the level of credibility of the people who are making the claims, but really credible people have been blowing the whistle for decades. So this makes me wonder if he's been given some inside information.

This morning I woke up to headline where Dr. Phil says that he has been given insider info and that "our government has been lying to us for 79 years", raising my suspicion even further.

When you zoom out and look at how all of these individual gears and cogs in the government are slowly moving in the same direction it starts to look like we might be seeing a coordinated government disclosure of...something.

The alternative is that we have a truly epic story of waste, fraud, and abuse on our hands. One where some clever clogs in the military figured out that the best way to steal money from the government was to claim to be studying UFO technology, and that they somehow got away with it for 80 years.

I'll be continuing to follow this closely as it unfolds, until next time, watch this space.

It has been a journey, sparked by a conversation between my wife and I about how insecure the system one of our homeschool co-ops uses for registration and I said (with much bravado), "I bet I can do a better job than that in a month!".

Checking my notes and git commits, it's been almost exactly a year since the initial commit (July 14th 2025) so I'm right there, in the ballpark 😊

The standalone product, an open source student registration system was technically built in about 2 or 3 months, but access to increasingly powerful large language models tempted me to try to turn it into an enterprise level product, and that's where the rest of the time went.

There is a lot of talk online about whether AI is really a game changer or not, and to me the proof is in the pudding. While most folks have been asking the question and wondering, some of us have been quietly building. That's really only one way to answer the question.

I can say without any doubt that ruff registrar as a product would only exist in my head as something nice to do "when I have some free time". At best I would have completed the community edition and shared it with some friends, maybe step in to replace the legacy app I mentioned earlier, but I definitely never would have had the time to launch it as a SaaS offering.

The biggest takeaway I learned from this experience of going from idea to launching a product is the amount of time and effort that goes into it. I used to think that me not being able to launch any of my solo efforts in the past was purely a skill issue, but now I understand that I had the skills all along.

I just didn't appreciate how much work it would be, I was looking at the foothills and finding those daunting without even seeing the mountain peaks soaring behind them.

My last refuge was the news podcast feature built into Android OS, you could select from a lot of different news sources and hear their morning headline readings to get a good sense of what is going on in the world. A couple of days ago that feature stopped working and the only thing Google News will give me is an AI readout of the days headlines, it's not the same and it caught me off guard. In hindsight I shouldn't be surprised, google does love to kill a good product.

I'm really not sure what it will take to set up a similar experience, but ironically I'm tempted to use AI to re-write the podcast news feed the old fashioned way.

It's early summer so my garden is humming, most of the summer bloomers are pushing petals out now and I've been spending a lot of time scratching in the dirt and preparing for their arrival.

I've also been working on ruffregistrar.com and that is coming along very nicely, I'm close to launch and getting tiny butterflies because I have real skin in the game. Every second the site runs costs me and it may take a while to get my first paying customers, so until it really launches it has been taking a lot of the rest of my free time. Meanwhile I've been working with one of the co-ops my kids are in to start alpha testing things to make sure they work.

Real work at seekr has also been busy, but fun. I don't want to talk too much about what I work on, but I'm on the platform engineering team which means a lot of my work touches every engineer in the company.

The Knicks are NBA champions! 👀

I'm an extremely lapsed basketball fan, mostly thanks to growing up a fan of the Knicks and the 76ers, I can't be blamed for checking out after the Marbury and Carmelo Anthony eras. I'm happy for them though.

We have all the ingredients we need for an amazing summer, but the political and economic tensions continue to provide a bleak backdrop of storm clouds off in the distance. The US/Israel war with Iran seems all but certain to go on indefinitely in a truly Orwellian fashion where we are always in a ceasefire and "days away" from a peace deal, but we're still attacking them, they are still attacking us, and the Strait of Hormuz is "open", but only if you're friendly with Iran and willing to pay a toll, or you'd rather roll the dice with an American Navy escort.

I've always been a close follower of the news, and never in my life have I ever been more tempted than right now to just tune it all out. But then I remember all of the people who are suffering at the hands of my government, and I realize that the least I can do is not look away.

I find myself shocked that after 10+ years of Trumps constant lying that now the entire worlds economy is being propped up by the idea that maybe this time he is telling the truth.

It can't last though, the Strait remains closed, the ceasefire remains a ceasefire in name only as Israel continues to brazenly annex Lebanon, flattening every square inch of housing so that no one can return. They aren't even bothering to make up excuses, they don't have to because the crisis they duped Trump into manufacturing is sucking all of the air out of the room and providing Israel with all of the cover it needs.

I'm worried about what happens next.

A lot of the economic damage has already been done, we can't plant fields now that were left unplanted due to the high cost of fertilizer. If the Strait opens tomorrow tankers won't magically appear in American ports. I'm sure you're reading the same headlines that I am.

Kitchen table conversations at my house lately keep coming back to whether we should plan our summer vacation knowing full well that by early July gas might be $7 a gallon. Or will we be glad we held onto the cash during the coming economic storm? Will this be the last vacation we can afford for a couple of years?

I'm sure we're not the only family having this debate, booking late, quietly impacting the economy downstream from us. The little cottage we've been staying at for the past three summers had maybe just gotten to the point where they could count on seeing us, now maybe they are delaying expenditures of their own.

There's no taking any of that back, it's no longer a question of if it's going to hurt, people are already feeling the pain. The only questions left are how much it's going to hurt, and for how long.

Now, in my defense, the scope back then was much smaller. I was going to self host my home grown student management system, a more secure version of the registration system a co-op my family is a member of uses now. That old system was donated by some previous parent who was a a programmer, but by this time that family was long gone, their kids are graduated, but they still support the website out of the goodness of their hearts.

I planned to use the old site as a spec and rebuild it using modern tech, I did some perfunctory searching and didn't see any open source tools that did what I wanted and I dove in.

I faced the usual challenges of agentic coding, I'm sure I talked about those in detail in my previous post, but towards the end I started to bump up some more subtle problems that I knew existed but didn't realize how annoying it could be until I really experienced it first hand.

First, as the project gets bigger and more complex it becomes more difficult to get Claude or Codex to stick to a problem solving style. Most programmers have a handful of patterns that they like to apply, and generally you'll see problems throughout a project that is maintained by one person or a very small team, that you see the same patterns applied to various problems throughout the codebase.

LLMs know all of the patterns though so they tend to use them all and sometimes they will use one pattern to solve a problem in one place, and then an entirely different one to solve a nearly identical problem elsewhere.

The other issue I ran into this past month was LLMs weird inability to say "No", or "That's a bad idea".

It seems that even when provided with very large context windows LLMs are still lacking the ability to understand the goals and overlay that on top of the context window so that each iteration pulls us closer to the solution. Since LLMs lack that, working with them feels like pulling a random, very talented, CS student off of the street to work with every couple of hours. Just as it feels like they are beginning to understand the project you have to dismiss them and grab another random student and start fresh.

I'm going to say upfront that the following example is, in hindsight, on me. But I asked the most advanced agent I have access to at the moment (CODEX 5.5) to basically solve one of those triangle problems where you can only have 2 of your 3 needs met. You want the job done, right, fast, and cheap, you have to pick two. Agents don't really understand that paradox and will happily burn a lot of tokens trying to oblige. Eventually it will present the problem as solved, having only actually done 2 of the 3 things and when you take a look at the code for the 3rd requirement you will inevitably find some kind of awful hack job.

When I was a very junior developer I used to get myself into messes like this all of the time, but as I gained more experience I learned to spot problems shaped like this and avoid the pitfall. The only way to really spot these problems though is to fully understand how your tools work, what their limitations are, side effects, etc. This is something that can't really be handwaved away by a general purpose coding agent.

At the risk of repeating myself, building complex things is still hard.

I grew up in a religious home, mostly Christians (Dad's side) and Muslims (Mom converted and remarried), all firm believers in God. I have to credit my mothers midlife conversion for planting the seed that eventually blossomed into atheism and agnosticism. It is very easy to indoctrinate a child into the culturally dominant religion, but it's a lot more difficult than you would imagine to convert a child who was in the process of building a logical framework on top of their foundational religious beliefs.

It obviously leads to a lot of questions, children are naturally logical and I was no different. I ended up rejecting religion so strongly that, without realizing it, I self indoctrinated myself into the materialistic world view. Only now am I starting to realize that I may have overcorrected, science doesn't have all of the answers and there are some rather large elephants in the room that science tends to ignore.

For me, the biggest problem with a purely materialist worldview where we're all meat robots and our conscious experience is simply an illusion, a byproduct of neurons firing off in our brains, is that it takes away our free will. It means our actions can be predicted deterministically by parameters outside of our immediate control.

This leads to my second issue: that in a universe without consciousness or free will there is also, fundamentally, no meaning. It's all just a happy accident, right down to the fact that we happen to be experiencing it, pure coincidence.

I don't know about you, but I find it a little suspicious that a purely materialistic world view robs us of our free will just a surely as an omnipotent and omniscient deity does.

A conscious universe represents another philosophical path where, by accepting that conscious experience is fundamental I can regain my free will and inject meaning back into my existence. Instead of feeling that on the timescale of the universe, my 40 years of life are meaningless, as a conscious being whose purpose is experience conscious awareness, then every millisecond that I'm awake matters.

The trade off is that I'm able to choose to accept that the universe is not completely deterministic, and that we can't know in advance what the universe is evolving into.

I think I'm okay with that.

If you ignore the news you could be forgiven for thinking that, outside of the recent war and subsequent gas price increases, things are pretty good. As I putter about in my garden this spring I find myself focusing inwards on this small patch of space that I can exercise control over. In those moments digging in the dirt I understand the temptation to put our heads into the sand.

War? What war? There is no war in Ba Sing Se my garden.

I guess I have this fantasy where, no matter how bad things get, if I just keep taking care of the garden that I'll be able to keep my house, and my job, and hold my life in place while the rest of the world spins out of control around us. By digging my hands into the dirt I'm trying to put down roots and bind myself to this spot, this place in time.

When we moved in I found it easy to picture the next 30 years in this house, now, only a decade later, it all feels so precarious.

The talking heads are saying that we're a week or two away from the worst fuel crisis in the history of the world. I don't think anyone in my age group can appreciate the significance of that because we have no working memory of the last fuel crisis in the 70s. I've seen pictures of the lines at empty gas stations but even during the worst recessions of my lifetime, we were always able to get gas, even if many of us couldn't afford it.

Right now we're all still driving around using fuel that was delivered before the strait was closed, and even in a best case scenario where the ceasefire hadn't failed, it would have taken weeks to bring things back online, so there was going to be pain one way or another, the only question was how bad it was going to be.

Small tangent but worth noting here, China is sitting on massive oil reserves, and they have the greenest energy grid, they will remain largely unaffected by this crisis, so put in a pin in that.

Everyone else though, now that the ceasefire has most certainly failed all we know is that we're careening towards the edge of a cliff, and the US leadership is still operating on vibes and a strong sense of invincibility.

What am I supposed to do with that information? Fill my spare gas can?

Rather than try to plan for that it's easier to just plant another tomato plant, pull more weeds, and pretend that nothing else matters.

So, how's your garden doing this spring? My basil is already popping up and it smells like heaven.

Since this most recent firing, Trump spent a week threatening to end the Iranian civilization, if they don't "Open the fucking strait, ya bastards". I had to log on to truthsocial.com for the first time in my life to verify that he really said that...

Just before the (taco) Tuesday deadline that he set for their annihilation, he decided to accept a ceasefire deal that the Iranians had sent over a week earlier. The next day Israel violated the ceasefire agreement by unleashing a massive coordinated 10 minute attack across many parts of Lebanon and especially Beirut. They then claimed that Lebanon was never part of the deal, the US agreed, and just like that the ceasefire was over practically before it started.

Ed Note: At the time of this post the ceasefire is technically still in effect, but the US is also now blockading the Strait of Hormuz, another clear violation.

All of it is exhausting, conspiratorially I wonder if maybe that's the point? Rationally I think that's giving these fools too much credit. No, there's no 5th dimensional game of chess being played here.

What we're watching is just the random gyrations of selfish, narcissistic, self-absorbed, assholes with too much power and an inability to self reflect being enabled by a mass of citizens who think that these are good qualities to have, actually.

They are going to break the world trying to mold it into their warped vision for the future of humanity, and then, if we let them, they'll blame someone else and walk away, leaving the mess for future generations.

We shouldn't let that happen, they are committing war crimes, and they should all be held to account.

The most notable of these is a new garden dashboard page. I thought it would be nice to know for sure when the seasons changed, and what to plant when, so I used the weather data to build a monitoring system that will keep track of the number of warm/cold days and nights so that when a certain threshold is reached we can officially declare the start of a new season.

I also included a database of all of the plants I grow, and the seed packet data will be used to automatically notify me when it's safe to sow a given seed, or when to transplant, etc.

Eventually I'll work my way back to the ruff-registrar project, I left it in a pretty good spot, but I tried to launch it to production pre-maturely and got burned by an Azure bill I wasn't expecting so I haven't looked at it after spending a couple of days scrambling to try to shut everything down.

I have a spending limit in place so it wasn't the amount that hurt, the issue is that once I hit the limit all of my other sites got turned off, lesson learned.

In other news, the Yankees are off to a decent start. They just opened the season with a 5-1 trip out west, it's been a while since they've done that, in recent years they've been happy to get out with a split. Opening day in the Bronx is tomorrow and I'm excited for that.

Summer is my favorite season but the way spring heralds all of my favorite things to come makes is a very easy close second. I'm going to get back into the dirt while there's still a bit of sun and warmth left and try to soak in as much as I can.

I feel bad centering myself at all, as an American I know I'm responsible for the actions of my president. I know that, for the most part, I'll be insulated from the levels of pain that are going to be felt across the globe. My anger that the people in charge the bombing campaign are demonstrably incompetent ignoramuses obviously is nothing compared to the anger of the people whose home are being bombed by those same idiots.

Today after a 5 week campaign where no real objectives were declared, just incoherent rambling, Trump has declared victory and says that the closed Straight of Hormuz isn't really his problem, America never needed that oil anyway, and if other countries need the oil they "can just go in and take it".

I know I keep repeating this, but even though I knew that America was on the decline and I thought I understood what that meant in academic terms. I always imagined that it would be a slow, inexorable decline, this feels like we're going over a cliff.

Trump and his Department of Defense, all staunch GOP members who in saner times would bitch and moan about every penny added to the deficit are blowing hundreds of billions of dollars per week on this war that is getting us absolutely nothing in return. Just lighting money on fire with no plan, no victory conditions, just fucking vibes man.

I feel equal parts ashamed, enraged, and dumbfounded that someone this stupid managed to take over our country.

I'm also afraid, I'm very worried about what's going to happen next.

I think that the reason I have such a hard time finishing things, even with the help of these powerful agents, comes back to the old 80/20 problem, where the first 80% of the work takes 20% of the time, and the last 20% takes the remaining 80%. LLMs haven't solved the last mile problem, not yet anyway, the unintended side effect is that the cognitive load on the human developer is dramatically increased.

In my experience the best way to develop software is to make something very, very small, that works, and then you iterate on top of that. As you do so the kernel of the way the code works becomes cemented in the developers mind slowly over time beneath the layers that are built on top of it. This helps me develop a good mental model of how the code works, and agentic coding short circuits this in a couple of ways.

First, they tend to generate way too much code to start with, and it's getting worse as context windows get larger. Second, they have no problem re-writing whole swathes of code from scratch to solve a problem, which prevents me from developing a good mental model of the code, so I frequently have to ask the agent to explain to me what is going on.

This reminds me of the quote

Complexity can't be eliminated, it can only be moved.

I think this is true in the case of agentic development as well, we've sort of taken the complexity out of writing code and transformed it into a different kind of complexity. Where I used to have a small number of projects that I knew well enough to jump back into at any time, but was blocked for lack of intelligence. Now I've unblocked all of those projects, plus added a half dozen more, except I don't "know" the codebases of any of them.

So while before I didn't want to work on a project because of whatever known issue was blocking me, now I don't want to work on projects because I'm low key dreading the process of re-mapping their complex code bases into a mental model in my mind.

Sometimes too, I feel like watching the agent burn through tokens while it's "thinking" is just as frustrating as banging my head against the wall myself, just frustrating in a different way. Instead of whitling away at the problem and slowly reducing frustration by increasing understanding until the problem is solved, now I watch an agent re-write a couple of classes before prompting it with something like, "the code couldn't have been that wrong before, could it?" and getting a response like, "you're right, this could be done more concisely" and changing 10 lines instead.

So what's the verdict?

Even after saying all of that I don't think humans, generally speaking, will ever write code again, I'm still going to continue generating all of my code as well. But I wanted to report back from the field and give a realistic view of what LLMs are capable of. Even in my day job, I spend a lot of my time applying the latest models to a variety of problems, this week Codex 5.4 dropped, and I've never felt more secure in my role as a developer.

One part of the argument is that whatever your produce will simply be consumed by the machine. The other part is that the machine will then generate so many copies that your original will never be found, so why bother?

I think that humans will always value stuff made by other humans because we are all connected to each other as conscious beings in a way that we are not connected with rocks and machines, not yet anyway.

Philosophically I find myself leaning towards dualism and panpsychism, I think there's more to reality than just the "stuff" around us, and I suspect that consciousness is fundamental and I can't draw a line in the sand where I would say, "at this level things are too simple to experience consciousness".

So I say all of that to say that conscious creations come with something extra, maybe it's simply the ability to "re-train" our own mental models on the fly. I can't prove it, but I have a very strong feeling that the words I'm writing right now aren't being driven by a non-deterministic algorithm bouncing through a matrix of weights, even if those weights are constantly being updated as each Planck second of my life ticks by.

I think it's our ability to experience consciousness that allows our models to continually update and generate new thoughts and ideas, or put them together in novel ways.

That's a very long and complicated way of saying that conscious beings have the ability to experience and enjoy their own creations, and tell you whether they think its subjectively good, in their experience or not.

In my opinion this puts the ability to create stuff that is uniquely new and good solely into the domain of conscious beings, because if you force the model to only generate "new" stuff, it will be effectively random because if you ask it to write 100 songs and pick the 10 it likes the most, it can't.

Meanwhile, actual good new content blessed as such by a conscious mind will become more and more rare, and simple supply and demand tells us what happens next.

Finally, I don't mind being consumed by the various models of the world as they inch their way towards possible consciousness, in fact I'm honored to breath my own soul into their corpus. If we are giving birth to a new form of life, my little blog posts will live forever, my own little ghost amongst many others in the machine.

Meanwhile, in the real world, we're like

• Hand out as many copies of the latest models to as many people as possible
• Make sure the models have as much autonomy as possible, we don't want to get in the way of the model
• Let's use the model for military purposes, with no restrictions

Are we mad?