Lynne – Lynne's Take on Tech

Thoughtful Thursday: AI IPR Slight-of-Hand

Since the halcyon days of Xerox, people have copied other people’s works. Many a student in high school, facing the daunting task of a five page research paper, has painfully copied from the Encyclopedia Brittanica, sans quotes, hoping to pad out the white sheets with ideas not their own.

Open source is no stranger to this practice. It was considered radical to actually keep provenance of code. Often when someone needed to write, say, a new driver, a working standard driver would be copied and then modified with the new hardware-specific functions. William and I even wrote an essay entitled “The Life Cycle of a Driver” which looked at the code history of a driver from the original author in the old Berkeley Unix PDP-11 days to it’s eventual inclusion in 386BSD on the X86.

Open source made it very easy to copy or “fork” entire OS releases. The only control on theft was a little copyright notice that allowed modifications and use as long as the author’s names and provenance were maintained. This small request lasted about two seconds in the OS wars. So often we found novel work we and others we knew absolutely had done that was now claimed by a new set of people who probably didn’t get beyond cp in the manual. William called them “tyros”.

William and I viewed an operating system’s collected code works more as an anthology or a collection of short stories than a singular work. As such, for 386BSD in particular, we viewed ourselves as editors, selecting the content from short works such as a driver, modifying it as needed to function, and putting it into place. Sometimes we added names to the copyright when it was substantially revised, maintaining provenance and placing new author’s names next to the older names as the original content is altered. In this case, one could see the changes to the code as it passed through different hands. Other times, when the modifications were minor, we left the names as is, considering the changes simple copy editing to fit the work into the release.

The distribution itself, the “book” so to speak, was considered a unique work. Our OS anthology release was entitled “386BSD by William and Lynne Jolitz” as the editors of the anthology and also authors of unique stories, such as role-based security (see Source Code Secrets: The Basic Kernel, Appendix B: A Blueprint for Role-Based Network-Level Security for further details). We always kept to that discipline, despite some grumbling from the tyro forkers.

Even Linus Torvalds remained credited as a 386BSD contributor for an early floating point emulator. While William had done the port with hardware floating point, Linus had to create an emulator for the 386 because the hardware was out of his budget. I think with all the licensing of Linux from big companies, that is no longer a problem for him. Andy Tenenbaum, the creator of Minix, from which Linus took much of the original Linux system, is free to chime in here.

So how does AI fit into this meditation?

AI is an interesting tool. It promises to automate, investigate, and discover new approaches mined out of complex datasets. And it has to some extent, as recent advances in drug discovery and mathematical proofs demonstrates. However, these research projects are based on very carefully designed datasets. Dirty datasets tend towards misleading results. When people can die, it should matter where the ideas came from, especially if it’s not monitored and proved by an experienced and truthful person.

Public AI has a very different agenda. Sucking up all the books, art, music, and essays from creative people since time immemorial appears like a great idea. It’ll be just like a person, right? We can create an AI even better than Picasso or Mozart! We will see “new” creations! Look at the hype! See how excited you must be!

The reality is AI is a tool. Just a tool. It can only spit out what it is fed.

But AI can do something for a tyro that the student copying out of an encyclopedia or a person using a cp command must do themselves: It quietly removes the authorship and provenance.

As the recent ruling out of Germany over Suno’s AI-generated music demonstrates, infringement is widespread, open, and blatant. The examples of songs in the proceeding were devastating. By just listing genre and some lyrics, Suno would generate songs that sounded to an ordinary listener identical to a pre-existing song, right down to beat and the sound of the original artist’s voice. “What was missing”, you might ask? The original artist’s credit, that’s what. No credit. No royalties. Screw you, poor artist.

The monetization strategy of feeding the AI maw is now quite clear. Companies like Suno, enabled by the Big Guys like Google, are structured to detach artist’s credits from their works and move the body of those works to company-owned and royalty-generating “new” works. The company now owns the IPR, enhancing its value to shareholders. The original artist gets nothing.

In plain simple English, companies are moving the large body of creative works to company-owned and edited works. This tech slight-of-hand will make them much much richer. And everyone else much much poorer.

This isn’t the end, however. The next stage in this approach is to make searching and finding the original artist an impossible task so as to obscure provenance further and reinforce IPR relabeling. Already, small artists, writers, and musicians are finding they are no longer easily findable by consumers. Google’s entire AI monetization strategy is to drown search by generating their own summaries so people stay with Google and click on their paid ads rather than click to another website not under their control. And it’s a quick step, no more, to completely eliminating dissenting or opposing views as if they never existed.

I read George Orwell’s 1984 when I was heading to UC Berkeley as a freshman in 1979, and the lines about generated songs for the masses always stayed with me:

“The tune had been haunting London for weeks past. It was one of countless similar songs published for the benefit of the proles by a sub-section of the Music Department. The words of these songs were composed without any human intervention whatever on an instrument known as a versificator. But the woman sang so tunefully as to turn the dreadful rubbish into an almost pleasant sound.“

We all know how 1984 ends. Are you ready to love Big Tech like Winston learned to love Big Brother? Because it’s coming faster than you might imagine.

AI, Vibes, and my New Office

I’ve taken some time away from the industry to move to a lovely new office. But I’m still in Los Gatos. So if you’d like to meet for coffee, hear some startup stories, or just chat about the industry, drop me a line.

The inexorable AI march continues of course, with deep-think execs surprised! at the real cost. Frankly, I’m surprised that they’re surprised. The entire revenue pitch for AI has always been a subscription and per-use model that gets them hooked and keeps them captive. I guess they were so obsessed with how the quarterly numbers were going to look so sexy without all those pesky engineers that they just didn’t think that maybe, just maybe, that free stuff wouldn’t stay free forever.

Ditto for “vibe programming”, the snakiest of snake oil imaginable. Yes, of course a good engineer or software programmer can and should use AI to make their work more efficient. They have the expertise to know how to streamline their work, and they know what questions to ask and what to ignore. Knowing the right question and knowing where to apply the work is how experts earn their pay. Ignorant bluster isn’t a substitute for knowledgable analysis.

We are in the age of confidence men and women, selling the rubes on what they want to hear. And no one is immune to its siren song of self-aggrandizement and greed.

AI is a fine tool. Like a sculptor’s chisel, AI can deftly shape the code. But the vision and architecture is in the mind of the creator. Only the sculptor sees the statue in the marble before the first touch of the chisel. And only the engineer sees the true form and architecture of the code and the ultimate product before tools are applied.

Let us all use our AI chisels wisely and well.

Sedate Sunday: AI Impacts On Software Developers and that Pesky Recession

Millions of Americans marched in yesterday’s No Kings rallies, although you wouldn’t know it from the controlled state press. Heck, i’ve had to go to places like reddit to get a feel for what’s going on with ordinary citizens. The US is at war, but nobody wants to call it a war, except perhaps the unfortunate young people in the military sent on a benighted mission to take some waterway we really don’t need with weapons that are inadequate to the task. And as the stock market wobbles now that the Strait of Hormuz is closed by Iran in retaliation for the US/Israel strikes on the country, the price of a gallon of gas, along with some spring travel, has gone straight up the flue.

But it’s a lovely Sunday in Silicon Valley, balmy and sunny. A perfect time to reflect on how jobs will be impacted by AI.

AI is not a “new thing”, and it’s not always a “bad thing”. AI is an excellent tool to analyze and respond to complicated contradictory datasets.

William and I had planned to use AI for 386BSD. An operating system and written works supported by a couple of enthusiastic people is a difficult prospect to say the least. We not only had extensive source code repositories — we also had extensive often unpublished writings on those very code bases. And it was impossible to respond to inquiries timely about this project. It vexed us.

William’s untimely death prevented us from taking action in this area. But that doesn’t mean it is yet impossible. New and accessible AI tools to mine, collate, analyze, and respond are now available. Open source documentation, support, and updates would benefit from these tools.

While the constant refrain that “AI will take all the jobs” is frightening, it’s also not correct. But it grabs attention from dazed and scared people worried about their next paycheck. It sells papers, so to speak.

The most frequent press focus on AI is software development. This is because the people working on AI are software developers. We eat our own dog food, as the Microslop folk love to chant. We know about software development, so we think “Hey, let’s make it do software development”.

Silicon Valley also loves to sell out their own. They always have. Our love of our own meritocracy results in a flawed reductionist nostalgia. Beginnings are difficult things, and we often start and stop before we achieve any success. We usually fail, but even the failures can have meaning. These lessons are forgotten once a startup becomes a success, sadly.

So the developers and inventors and folks who did the hard work are forgotten. Resentment of those “costly” developers festers. And then the Silicon Valley C-suite says “Hey, let’s use [pick a tool] to get rid of them. There’s a whole lot of code on the web. There’s stack overflow. Let’s raid it. Who needs those guys!”. Hence, the current fad.

Many Silicon Valley companies overhired in the last decade, mostly to fill offices in their corporate real estate kingdoms. So reductions were expected. In addition, we’re in a late stage economic cycle before a recession. Holding on to lots of people developing products that can’t be sold in a recession is scary. Yes, the C-suite gets scared just like ordinary people. So they’re getting ahead of that event before it happens — often to excess.

This is not our first recession rodeo, folks. Been through it many times here in Silicon Valley. We got startup funding several times during the darkest recession times. Why? Because trying new approaches is a time-honored way of getting to the stage after the recession, which is growth. That’s where the good engineers and inventors and software developers will be. You can’t automate innovation.

So now, who is actually at risk? The answer is anyplace where costs dominate or innovation stagnates.

Companies and even individuals first reduce costs by looking at non-critical areas. Marketing, PR, and social media. HR. Travel. Creatives. They’re on the cutting board, as always. AI is already chatting with customers, and sometimes even solving the problem. Sometimes.

Stagnating innovation is a bigger issue. Product development at companies will be in a free fall. People will be fired. AI will be blamed. But it’s just the same old cost-cutting in a new coat. Gotta blame something, and blaming AI is easier than admitting your company is not going to do anything new for a good while. So batten down the hatches, folks. AI isn’t necessarily coming for your jobs. It’s just the convenient excuse.

The dark and looming monster coming for your jobs is the anticipation of a recession. Will it actually occur? I don’t know. I don’t know when or how or why. Nobody does.

But the fear-of-losing-out culture is now the fear-of-moving-too-late culture. AI is the excuse, not the cause. Look behind the corporate curtain. Think critically. Read. And plan for your future. Because no one else cares about you like you do.

Thoughtful Thursday: Data Center Marauders and SV Villains

Today as our country spins further into self-induced insanity, I thought I’d talk a bit about (drumroll) data centers! This is ordinarily a rather dull topic to most, consisting of equipment and deployments and connectivity, hardware and software and systems, processors and networks, and of course, operating systems.

Since it is a dull but necessary topic that keeps things running, I could not resist writing about them back in the day. I even put together a book for Wiley entitled Inside the Internet Data Center in the early 2000s which predicted many Internet issues from the perspective of the data center administrator and how to deal with them. It was an enjoyable exercise. Unfortunately, the publisher decided it was an uninteresting field of study and canceled the book. I got to keep the advance and soon moved on to other projects. But data centers will always hold a fond place in my technology heart.

Now, everyone is talking about data centers. But not fondly. Not as an intellectual exercise. Nope, they are talking about them as technological marauders, stealing power and water and land from ordinary citizens in small towns. These entities are placed without any consultation or approval of the citizens in their communities. And the deals for taking the land and power and water are too often under nondisclosure, meaning nobody knows until it’s too late.

The big players in AI – Nvidia, Meta, Google, Amazon, OpenAI, and Microsoft – are now seen as the faces of villains. It’s a PR disaster for Silicon Valley. And yet, this is the endgame move. There are lots of pieces and players – and billions of dollars involved.

At the core, a data center is simply a processor receiving a request, accessing a database entry, and sending a result back. The devil is in the details. And there are a lot of devils, ranging from the actual hardware management and maintenance, to connectivity requirements, to database management, update and integrity — not to mention the software and operating systems that superintend these functions.

These operations require power. Lots of power. And they produce heat. Lots of heat. Hence the massive power and water requirements. And the increased demands of AI mean even more power and water for those AI chatterboxes telling folks that they can do no wrong, even though the actual interactivity to the end user is essentially meaningless.

According to McKinsey, almost $7 trillion will be invested in data center acquisition and development by 2030, $3 trillion in real estate and power infrastructure and $4 trillion in “computing hardware infrastructure” ($3.5T in servers and $.8T in storage). This is a mind-boggling number of processors.

Nvidia as the leading vendor of GPUs is beyond excited. At GTC this week, Nvida CEO Jensen Huang hailed these numbers, claiming he sees “at least $1 trillion” in revenue from Nvidia products through 2027. “I am certain computing demand will be much higher than that.” As Motley Fool noted, “This is startling — in a positive way — as it comes only five months after his earlier forecast, suggesting enormous momentum in orders in recent months. Also, Huang’s words imply that, from what he’s seen so far, he thinks revenue could surpass the level of $1 trillion.” If so, McKinsey’s predicted spend is underestimated.“

The real estate side of the equation is equally compelling. Data center REITs are all the rage. While partners in these ventures are often carefully shielded from local scrutiny, huge funds are usually involved in the acquisition of land and construction of essentially a bunch of big concrete shells. Rural locations are targeted. Land is cheaper in rural locations. Political resistance is diluted. And oftentimes water available for agriculture can be recaptured for equipment cooling. In addition, it is easy to buy a few local officials (and relatively cheap) and obtain through strong-arm tactics power easements to the datacenter.

When large corporations come into a community, they usually attempt to co-opt it by selling folks on jobs. However, once the datacenter shells are put together, there really isn’t much of a need for unskilled or semi-skilled labor. A large data center requires very few on-site personnel. Most software and systems administration is done remotely by engineers. Hardware failover means that a lot of CPUs and GPUs can burn out without having to swap out the damaged equipment.

The modern data center isn’t a factory with people working in shifts and products trucked to customers. It’s a factory for analyzing your most personal secrets, foibles, and thoughts, circumscribing every purchase you make and every comment you write, and monetizing it.

If fund managers believe that $4T in real estate is worthwhile, think of what value they place on your personal data. To get a 10-fold investment return, youre looking at the data value of literally hundreds of trillions of dollars estimated. These numbers rival the GDPs of major countries and blocks, with the US, China and the European Union in the $20 trillion economy group. That’s two countries and one economic block.

Now, these numbers are incredible, but also quite ephemeral. The same value of data of citizens incorporated implicitly in GDP as a complex factor is often claimed as a pure monetary value by AI enthusiasts. Nope, this isn’t the ROI. It’s about controlling the mindspace of investors and securing current investments.

What is real and tangible is the resistance of data center construction in communities throughout the US. And that antipathy has gone hand-in-hand with increased public fear of AI used for government surveillance. And most certainly some of the biggest advocates for its use have been companies like Palantir. People don’t like being reminded they’re being watched, even though that’s what the “free Internet” has been all about.

AI advocates have touted AI as the means for companies to find and fire people because they’re no longer needed. Lots of people. Huge numbers of people. This doesn’t make folks worried about their next paycheck feel real good about it. Especially when they open up their power bill and find it’s gone way up because a datacenter in their area has raised costs for everyone.

And finally, the debacles of AI misinformation from overeager AI advisors have become a joke, and a mean one at that. AI for the masses is not a trusted source.

This is how the standing of Silicon Valley companies has plummeted — not in terms of the stock market, but in terms of their trustworthiness.

William and I discussed the benefits of AI and data centers many years ago. Those benefits haven’t changed. But the mishandling, greed, and sheer contempt of major players in Silicon Valley and large investment funds have frittered away the goodwill and obscured the benefits.

It’s too bad. I really do like data center design. I like the software. I like the systems. I like the hardware. I like the benefits of AI.

But I don’t like the companies and pundits. Nobody does.

Sunday Reflection: Democracy is dying and Silicon Valley is Silent and Complicit

Alex Pretti, VA ICU RN, murdered 24 January 2026 by masked federal agents

The first thing I thought when I saw Alex Pretti was “He had a kind face”.

The second thing I thought was “He was shot down in the street for helping a woman as a medic”. (My father-in-law was a corpsman in WWII.)

The third thing I thought was “So many videos. Everyone can see it for what it is – a cold-blooded execution of a US citizen”.

The fourth thing I thought was “Silicon Valley is silent and complicit”.

Let me tell you how I got there.

William and I had a singular goal in launching open source Berkeley Unix in 1990 — to allow people to not only deploy servers under their control, but also to control the source code on which they ran.

This meant no company, no government, no billionaire, could sabotage, interdict, or otherwise censor the people or keep them from exercising their Constitutional rights.

When we released 386BSD 0.0 in 1992, I heard from people all over the world. Universities and colleges, NGOs, entrepreneurs. You name it. Most of these folks could never afford to host their own server, much less own one themselves. It was liberating and spawned universal access to the Internet.

But the tides of capitalism are brutal. Over time, monopolies grew to control great swathes of our global intellectual capital.

Let me be clear. A free society cannot exist when it is captive to the whims of a handful of billionaires and criminals. Nor can it endure the unfettered unhinged hate propaganda on social media secretly fomented by foreign enemies like Russia (yes, Russia is an avowed enemy I’m sorry to say) which we damn well can track and eliminate but won’t for a few cents more profit!

Who controls the press and social media? Silicon Valley billionaires.

Who controls the government? Silicon Valley billionaires.

Who is steadfastly feeding Americans a diet of lies and deceit? Silicon Valley billionaires.

These are very same people who made their success off Silicon Valley’s free exchange of ideas and technical works. But now for narrow and parochial reasons, the truth is blurred and waved away.

Tim Cook watches a ICU RN assisting an injured woman gunned down in the streets and then parties with Trump and Melania.

Elon Musk’s DOGE tech goons steal the social security data of US citizens and face no Republican censure.

Jeff Bezo’s Washington Post reporter Hannah Natanson’s home is raided in a ginned up fishing expedition and he shrugs it off.

Just to name a few. There’s plenty more right here in Silicon Valley.

There is this simple-minded belief that Silicon Valley is above all this. That we are a noble meritocracy. That we need not concern ourselves with the nation.

But we cannot be silent. We are complicit. Without us, these criminals would be knuckle-dragging in parochial ponds.

We provide the infrastructure for their hate.

We provide the support for their organization.

We provide the technical expertise for their exploitation.

And we are wrong. Silence and complaisance goes against everything Silicon Valley has stood for.

Silicon Valley is a success precisely because we do not suppress knowledge. We run by rule of law and not rule of brutality. We openly discuss and engage in new ideas. We thrive because of democracy. No other country holds our record for innovation and success. It is the core of our being.

Silicon Valley is the Golden Goose of America.

And now a handful of people who have made fortunes here in our community are intent on destroying that Golden Goose, for no particular reason I can discern other than that they’re bored. As the old saying goes, “Idle hands are the devil’s workshop”.

Renee Good, poet and mom, murdered 7 January 2026 by masked ICE agents

So do we stay silent and complicit while masked goons kill poets and nurses?

Do we turn away when the truth is strangled?

Do we turn off our minds as our freedoms are tossed away?

Or do we speak out and firmly remind our colleagues in Silicon Valley that they will be on the wrong side of history, morality, and ultimately, truth if they do not act in support of democracy, rule of law, and freedom of thought?

250 years. That’s how long we’ve lived in a democracy. 250 hard fought years.

I’d hate to see Silicon Valley destroy it. We’re far better than that.

Fun Friday: AIs are the Very Model of the Silicon Valley Elite

A lovely day is arising in the Santa Cruz Mountains, and another Fun Friday.

Rebecca Solnit commented on her Facebook feed today that “…AI has come for me too”, with lots of AI posts using her name and mimicking an article from 2008 entitled Men Explain Things to Me to snatch attention, hits, and money. Whether it works or not, I can’t really say. Facebook has a lot of cash sitting around to pay off bot farms, I assume, in lieu of the appearance of activity.

But what I found interesting is Solnit’s reposting of some selections from that article I had forgotten. As a well-heeled tech idiot held her and her friend captive with his “intellectual” rehash of “the very important Muybridge book that came out this year”, which was Solnit’s book by the way, so I’m sure she knew of it, she continues:

“He was already telling me about the very important book–with that smug look I know so well in a man holding forth, eyes fixed on the fuzzy far horizon of his own authority… So, Mr. Very Important was going on smugly about this book I should have known when Sallie interrupted him to say, “That’s her book.” Or tried to interrupt him anyway. But he just continued on his way. She had to say, “That’s her book” three or four times before he finally took it in. And then, as if in a nineteenth-century novel, he went ashen. That I was indeed the author of the very important book it turned out he hadn’t read, just read about in the New York Times Book Review a few months earlier, so confused the neat categories into which his world was sorted that he was stunned speechless–for a moment, before he began holding forth again. Being women, we were politely out of earshot before we started laughing, and we’ve never really stopped.”

I’ve worked in Silicon Valley for many years. I’ve had my writing on operating systems explained to me many times by eager men enthused by “that OS article”. I’ve had my technology explained to me by eager investors who “really got it” (Narrator: They didn’t really get it). I’ve even had lawyers suggest that my patents weren’t really my patents and should be assigned to someone else they like better and for no other reason than that it would make it more “salable”. Note in the last case they damn well knew my name was on them, and also that patent law frowns on this sort of thing, but oh well. Remember kids, don’t let your attorneys pretend they’re businessmen – they’ll come a cropper every time.

So this is normal operating procedure in Silicon Valley and always has been.

But what I found most amusing is you could describe the current AI nonsense in exactly the same words. AIs are the very model of the Silicon Valley elite with too much money and too little curiosity.

AIs don’t read the books they talk about.
They insist on your attention under false pretenses.
They drone on and on, precluding interruption.
They mash facts and fiction into an incoherent babble of words.

The only difference between the two is sometimes one can embarrass the elite. Sometimes.

Evolutionary essentialists often claim our only purpose is to reproduce ourselves. Looks like Silicon Valley has found a new modern means to do so. We are the Innovators!

I’m going to read a book and enjoy the day. You should too.

Fun Friday: Generative AI, Infected LLMs and Breaking Tulips

It’s a sunny and pleasant day here in Silicon Valley. A perfect day to chat about the imperfections of AI and the oddities of Breaking Tulips.

The extinct Semper Augustus Tulip (Norton Simon Museum)

Several weeks ago, so long now that everyone has likely forgotten, MIT put out a business survey of generative AI effectiveness, and found that hardly anyone is happy, except the consultants. “Just 5% of AI pilot programs from enterprises contributed “rapid revenue acceleration,” while the majority stalled and offered little financial impact”.

Of course, companies like Nvidia took a brief hit and then promptly went back up again, proof of the resiliancy of this latest Silicon Valley AI Bubble.

Given the wackiness of our current business climate and government, it is no surprise that people are clinging to generative AI as a beacon in the looming economic darkness. There are few opportunities for launching other technology startups as a few huge AI companies continue to suck in the majority of investment dollars and press hype.

New technologies are also a very hard sell in a risk-adverse market. Customers will only buy from extremely well-funded “startups” or well-established companies — if they purchase anything at all. And with the ever-present fears of economic upheaval, nobody wants to issue a PO, unless it’s for a sure thing.

And that sure thing is, of course, generative AI! Why? Because all the press and buzz says it is.

Behind the curtain there is great concern about results. That is why companies like IBM are actually doing a bit better. The assumption is there will be a need for high priced consultants to make generative AI a positive and lucrative business success. This confidence in the consultant pipeline and the belief that generative AI will eventually “get there” is what’s keeping the investment bubble going.

Oddly enough, contrarians are saying generative AI is just another “tulip bubble”. The actual Netherlands tulip bubble occurred from the early 1600s to the 1630’s. After a long period of fascination and investment in tulips by the aristocracy, the merchant class got into the action in a mass of speculation and eventual catastrophic collapse. It’s one of those “case studies” economics students love to pontificate about. So yeah, everything is a “tulip bubble” to some folks.

I’d like to say right now that our AI bubble is not a tulip bubble. If you want to go pick flowers, try crypto.

But there is one little piece of trivia about the tulip bubble that I do find applicable, particularly to broad-based LLMs. And that is the investor fascination with Broken Tulips.

Broken Tulips are a stunning flower. Induced by a virus, tulips will unpredictably turn from a humdrum bland solid color to an exciting streaked flaming riot of colors. These tulips were sought after by investors and caught the attention of the masses. Everybody wanted one. Artists painted them. Investors purchased them at huge sums. They were the height of tulip mania.

But a virus riddled plant is not a strong plant. And so it was with these tulips. They all died out very quickly — and so did the investment. According to the Amsterdam Tulip Museum, “Over time, the virus weakens the bulb and inhibits proper reproduction. With each new generation, the bulb grows weaker and weaker, until it has no strength left to flower and withers away.“ Not a good bet for seasonal returns. And since tulips are supposed to bloom every year, a dead plant is a dead loss.

Which brings us to another recent study, in the long line of studies, relating to generative AI LLMs and “hallucinations”, in this case by researchers at OpenAI. In Why Language Models Hallucinate, Kalai, et al. argue that language models are essentially encouraged to “guess” if they don’t know the answer, leading to absurd and false statements. They also state that this problem is persistent and pervasive due to the training and evaluation processes inherent in the process.

Current LLM models are akin to breaking tulips. They’re infected with incorrect assumptions that make them appear smart and decisive. Like a breaking tulip, this confidence inspires the customer to want more and more. In an age of complexity, nothing is more seductive than a decisive answer. And when a model seems to know more than you do, you can stop worrying and thinking and just defer to it. Even if it’s wrong.

Hallucinations popping up again and again are not a good basis for business decisions. Hence the MIT survey results.

Like that lovely infected tulip, the infection is persistent and insidious — to the point the LLM model may become too damaged to rely upon. Kalai, et al. offer no satisfactory cure for this infection. They state, “This ‘epidemic’ of penalizing uncertain responses can only be addressed through a socio-technical mitigation: modifying the scoring of existing benchmarks that are misaligned but dominate leaderboards, rather than introducing additional hallucination evaluations. This change may steer the field toward more trustworthy AI systems.”

In other words, businesses should go back to the drawing board and create their own LLM results based on binary true/false statements, carefully run, so as to mitigate the attempts to provide any non-verifiable answers. Which is how earlier models were typically done.

When a solution is sold as “simple”, telling the customer they now must do a lot of complicated work to make it trustworthy isn’t the best answer. But it’s the only path forward for these customers who are betting their future on generative AI.

Generative AI is not a tulip bubble. But some aspects of it are bubble-like. Given the promise it offers to business, I expect to see curated trustworthy business LLMs which are held proprietary.

However, for those heavily infected models used by the public, we can continue expect nonsense spewed out confidently and wrongly. I doubt most people will notice.

Sedate Sunday: AI Spending is Never Enough. Vine Revived?

It’s a pleasant morning here in the Santa Cruz mountains overlooking Silicon Valley and Monterey Bay. So what better time to think of AI and video?

According to the Guardian, our beloved tech overlords – Meta, Microsoft, Amazon and Google – have spent a whopping $155B on AI development this year alone. As the Guardian notes, this is “more than the US government has spent on education, training, employment and social services” the same period. Oddly enough, these are also the very areas where AI is expected to take over from all that expensive and pesky “skinware” (that’s people, in case you didn’t know). Such a coincidence.

Most of this money falls under the capital expenditures category, which is not an expense but an investment in infrastructure, R&D, semiconductors, and so forth, and hence runs under very different tax rules. Most of these investments I would say fall under the AI rubric only tangentially, but are categorized as “AI” to make investors happy. And happy they are. Big Tech (sans tariff goblins) have lots of money to spend on themselves.

From a Silicon Valley perspective, many of these investments are long-overdue, particularly in semiconductor design and datacenter platforms, along with requirements for low-latency low-loss networking. So this is actually a good thing for a broad swathe of hardware and software engineers and designers.

On a more amusing note, Musk announced yesterday that the Vine archive will be restored after seven years in cold storage after Twitter dumped it. I’d hate to be on the restore team. The bit rot coupled with mind rot would be formidable. But if you’re wishing you’d grabbed Aunt Martha’s six second blowing out the candles on her birthday cake over and over, now’s your chance, if it’s really still there. Industry and customers have long moved on to other video pastures, so this seems just another fair weather balloon soaring on bluster.

It’s still a beautiful day. Take a walk. Breathe the free air. Remember that nothing lasts forever. So take a chance today.

386BSD Release 0.1 Released on Bastille Day is Thirty-Three Years Old Today

I was thinking today about our creative work 386BSD Release 0.1 and the open source revolution William and I had the privilege to ignite on Bastille Day 1992, 33 years ago.

In March 1992, we released 386BSD 0.0, a fully open source operating system as documented over the prior two years in Dr. Dobbs Journal in our Porting Unix to the 386 series. 386BSD Release 0.0 was by design a very minimal release with novel code done by us filling the gaps. It had to be minimal, since there were concerns about proprietary code from old Unix releases of bygone years. By documenting and releasing this early OS as open source, we achieved a baseline release on which people could freely build.

And build they did! We received thousands of contributions, fixes and updates over the next three months. It was an immense challenge to inspect, vet, test, incorporate, and credit those responsible. But we did, with the support of the open source community, Berkeley Unix specialists, and Dr. Dobbs Journal editors. It was a remarkable time.

386BSD was the progenitor, platform, and inspiration of many open source projects. By porting it to the X86 processor, a proprietary academic testbed system only accessible to institutions and government became, in a moment, the means for any talented programmer to implement their ideas and expand their horizons.

A couple kids from Berkeley and a dream. That’s where it started.

As I sit and look at the current software ecosystem, I do wonder if what William and I achieved, with the further assistance of so many other wonderful people, would be possible today. I have my concerns and doubts.

But then I only need look at the tremendous value open source presents to the next generation of innovation to hope.

Keep the dream alive in your own life. It’s worth it.

A Tale of Two Universities

The wellspring of innovation in Silicon Valley is anchored by two major universities: the University of California at Berkeley (UCB) and Stanford University. UCB, ranked by US News in 2025 as the nation’s top public university and sixth worldwide, and Stanford University, a private university ranked as third worldwide, are both R1 research universities. R1 universities invest a substantial amount in research, award a significant number of research doctrates, and create the framework for advances in science.

These two universities are the powerhouses which drive Silicon Valley’s startup economy. One cannot understate their import in techology, medicine, aerospace, engineering, and science. Berkeley Unix, for example, was a major testbed for many operating systems and networking structures used today. 386BSD, which William and I spearheaded in the early 1990’s, pushed operating systems and networking into the mainstream and open source as the mechanism for new works.

This year UCB was rated second in space sciences by US News. However, the premiere MAVEN program based at UCB Space Sciences Laboratory and the University of Colorado Boulder Laboratory for Atmospheric Physics, like many other aerospace and science programs, was completely gutted in the most recent budget debacle, among other programs in aerospace. This is happening across every science, engineering, and technology program today in every R1 university!

MAVEN is a highly successful and well-published project which expands knowledge and technologies. But like so many other things we depend on, I suppose it’s not on TikTok enough to merit attention — hence its successes are ignored.

This reductive and unimaginative attitude is illustrative of how our current populist era sees the advances that Silicon Valley has created. We in the tech space have made the Internet possible. We have created the underlying hardware to transmit it. We create the economic incentives for startups. We are the information economy.

Yet we are treated terribly by a public and political system to which we have given so much. Perhaps too much. UCB, along with other R1 universities are facing catastrophic research cuts, layoff, and loss of top-tier scientists and researchers to international universities. Likely the next big advances will not be from the good old USA, but instead China and Europe. So much for America first.

To keep some critical research going and to keep commitments to their students, researchers and staff, UCB and Stanford may be forced to dip deeply into their endowments. UCB has an endowment overseen by the state of $22.6B. Stanford, in contrast, holds $37.6B in endowments.

The reason UCB has a substantially lower endowment despite its far greater research and teaching reach is due to the manner of how technology transfer is handled. UCB, as a public university, is torn between public responsibilities and the concerns of private technology locks held by companies and startups. This dichotomy has resulted in many different debacles, among them the Berkeley Unix lawsuits of the 1990’s and the CRISPR patent failures of the 2010’s. Simply put, UCB isn’t very good at monetizing their advances. Perhaps this is for the best.

Stanford, on the other hand, has no such public concern. Technology transfer is quite straightforward, and the easy movement of people between the lab and startups makes up much of the growth of their endowment. There are advantages to being private.

But despite the fact that Stanford (along with Harvard, Yale, and other private R1 universities) have also educated much of the current political, judicial, and financial ruling class today, these hefty private endowments have caught the populist eye and are ripe for plucking. A proposal to place a 21% tax on them has these universities quivering in fear. Stanford is already planning for budget cuts and layoffs.

The tech billionaires educated and nurtured by Stanford who are going to benefit the most from the latest federal budget do not appear to be concerned with the next generation of technology and innovation. Perhaps they are tired. Perhaps they are greedy. Perhaps they simply don’t want anyone to do better than they did, and they’re going to take it all to their grave — if they don’t find a way to live forever. It really doesn’t matter.

William and I gave back to UCB with our own sweat and blood and tears. 386BSD was a labor of love, and it changed the industry.

It’s a pity that Stanford, Harvard, and their ilk did not instill a similar sense of loyalty to Silicon Valley. Things might have been very different