M is for Midtown and Manhattan and Mamdani (and much more)

It’s been a year since I wrote about New York City and two years since I’ve visited. Who knows when I will be back. I do know I still love to read about it, which isn’t hard, since it’s often in the news. So here’s my annual post on NYC, brought to you by the letter “M”.

Midtown: First off, shout out to Midtown Manhattan, a much maligned part of the city. As this piece shows, it had quite the year, between the Knicks winning the NBA championship to Taylor Swift getting married. Mind you, it had some bad things happen, like this collapsing high rise,  That said, this quote sums things up: “Midtown is not usually cool, but, right now, it kind of is.” I’ve always had a soft spot for Midtown, so I am glad it is getting some love.

Relatedly, I really liked this piece: “As a child growing up in Midtown Manhattan, I learned to speak apartment very early.”

Mayor Mamdani: Much ink has been spilled on the new mayor of New York, the indefatigable Zohran Mamdani. It seems every day he and his team and promoting a new initiative for the city, from solving the city’s budget crisis to working on bringing cheaper grocery stores to certain boroughs. While he sometimes wanders into topics that don’t seem to make sense to me (e.g., the Koh-i-noor diamond debate) and while some of his goals may not be achievable (e.g., taxing the rich), I am impressed by his drive and his ambition.

I’m also particularly impressed by his communication skills. This interview he did with the New York Times is like a masterclass on how to handle difficult questions. I’ve seen him do this not just with people from the Times but other interviewers as well.

As for the Times, they don’t seem to know what to do with him, is my general thought. Going through this list of articles and op-eds about him in that paper, you get the sense that at best, they are highly doubtful of him. Let’s see how that relationship evolves.

Making do: affordability is a big topic not only in Mamdani’s New York but everywhere. The Times has a good series on how people with every day jobs somehow are making do in the Big Apple. The stories range from a family of 3 living on $500,000 on the Upper West Side to another family of 3 living off $25,000 in Queens. Quite the range financially and geographically.

Minimal living: speaking of affordable, visitors can always stay at these $100 hotels in New York City. Do you want to is the next question.

Maximal living: is this place affordable? No. Is it worth a look? As a fan of maximalism, I’d say yes.

Museum Living: not affordable or maximal is the Manhattan apartment of David Sedaris. In some ways it’s like seeing a museum or a nice art gallery. We should all take up humorous writing and live so well. 🙂
 

Martinis: if you are in New York and you’re looking for a place to get a New York happy meal, here are 13 essential spots according to Wallpaper*. And what is a New York happy meal, you might ask? Well according to the article: it consists of “an ice-cold martini and piping-hot French fries”. Sounds great to me.

Meals at 3 a.m.: I worried during the pandemic that New York might be backing away from being a place that never sleeps.That you might not be able to grab a bite at 3 a.m. like you once could. Well I worried for nothing. According to this piece by, there are still great places you can eat 24 hours a day. (And if want to know of more places to hit up, you’ll want to check out Fortney’s other pieces.)

The Metropolitan Opera: tough times for opera fans in New York (and those like me who come to the city on occasion to see opera, as New York’s Met Opera recently announces ‘necessary’ layoffs and pay cuts.

Massive: when I think of “massive” and “New York” last year, two things come to mind:

 

What to do on a day off

If you are Canadian, chances are you have the day off today. Maybe you have no great plans other than to hang out in a hammock or beach yourself somewhere. And that’s fine.

But if you are someone like me who likes to be productive even with down time**, then perhaps you’d like to not only do some reading but actually finish a book? Sound impossible?

According the The Guardian, you should be able to read these 20 brilliant books in a day. If you have an iPhone, you can download the book above right now and dig in. 

Enjoy!

** I know, it’s a problem. 🙂

On how OpenAI lost control of an AI Model that ended up attacking Hugging Face

On July 21st OpenAI published this piece about how they and Hugging Face partnered to address a security incident during model evaluation. Like many people, I was shocked by the news of how agents at OpenAI broke out of their secure environment and attacked the Hugging Face site.

Later, though, in reading other pieces, I was less focused on how impressive the agents were and how poor a job the OpenAI team did. Reading this piece in TIME, it appears the agentic software was left unmonitored on the weekend. That was bad enough. Worse, the secure environment was not air gapped in any way. Once the agents figured out how to break through any software barriers between them and the Internet, they were free to wreck havoc on the outside world.

As the San Francisco Examiner states: “A recent hacking incident involving a pair of OpenAI’s artificial-intelligence models highlights the cybersecurity risks of the technology — and a serious security lapse by the San Francisco company, computer-security experts say.” Indeed.

For more on how OpenAI screwed up, see the pieces linked to here.

P.S. It gets worse…I just came across this: OpenAI’s rogue models roamed the internet for 4 days and staged a second attack – POLITICO

On fires and bridges: Canada and Trump’s America

It is an odd time to be a Canadian, located north of Trump’s America. As in previous months, Trump has stepped up his attacks on his neighbors to the north. And as before, he’s slamming more tariffs on Canadian goods, this time due to the smoke from the forest fires, he says. Who knows? 

Here’s some links I’ve captured on the state of the disunion between the two countries:

All in all, a sad state of affairs. I am grateful for my good friends and neighbors south of the border. Let’s hope they prevail in upcoming elections. Let’s hope things get back to the way they were under Biden and Obama and Bush.

The odd bun flight about The Odyssey, the Sea People, and the Bronze Age Collapse


It’s an odd thing, what Emily Wilson did in writing this piece, An Uncomplicated Man, for the London Review of Books. Suddenly the discussion of Christopher Nolan’s new film turned into a debate on whether or not Nolan gets the history right. I want to delve into the second part for a bit.

It’s a tricky thing to criticize such things as the film’s portrayal of ancient Greece, since almost everyone concedes that you can’t really be accurate about history when it comes to Homer’s epic work, since it is not historically accurate in itself. So I was especially struck by the strong dismissal of Nolan’s references to the Sea People.  Such dismissal came at the end of the Wilson piece,  as well as this piece by a Historian of Ancient Greek (which is very good otherwise).

While not an expert, I would give Nolan a break on that, for a few reasons. One reason, as this piece explores, Did Odysseus Cause the Bronze Age Collapse?, is that it is hard to get a good gauge on who the Sea People were, so I am inclined to cut him some slack. A second and more important reason, to me, is that Nolan is using artistic license here. He is no more being historical when it comes to the Sea People than he is with certain costumes, and he sacrifices historicity for the sake of artistry. That’s how I believe he uses the reference.

That bun fight aside, I think it’s great that Nolan’s “The Odyssey” has fired up people’s opinions.  And it’s great that subject matter experts are weighing in on the film, the history, and the literature behind Homer’s work. More films should do that.

P.S. For more on the Sea People and the Bronze Age collapse, I highly recommend this book, “1177 B.C.: The Year Civilization Collapsed” by Eric H. Cline. It’s a fascinating story.

From Prada Space Suits to Mamdani’s kurta (Fashion links I am interested in, July 2026)

Fashion can be many things, from frivolous to functional, with stops in between. Here’s a dozen plus fashion links that center on the functional, the famous, the frivolous and more.

Functional
Famous
Frivolous?
Finally

A mouse, a house, and an inflatable chair (what I am interested in when it comes to design and interior design, 2026)

Thanks to the good folks at Yanko Design (and some others), I’ve pulled together a list of interesting links over the last year that touch on the topic of design (e.g., smart phones) and interior design (e.g. chairs).

Devices: Let me start off with devices. Some IT based, some not….

IKEA: Ikea is continually doing interesting work, year after year. I think these links show that:
Furniture: IKEA is not the only furniture maker with well designed furniture. For example…
Living Space: finally…

On the high cost of Canadian beef in the summer of 2026


Like many Canadians, I’ve been wondering why the price of beef is so expensive. This article from the CBC explains why.

To summarize the article: supply is low, demand remains strong, production costs are soaring, and the processing network is highly concentrated with inter‑provincial regulatory.

Canadian farmers need to growth their herd sizes, but it can’t happen quickly enough. Even as they are doing that, the demand for beef is not declining, despite costs. So who knows if larger herds will make a difference. Does it help that the price of fertilizer is going up due to a war? It does not. Do other factors uniquely Canadian make a difference? Yes they do.

Perhaps beef prices will decline in the long term. In the short term, you might have to be like me and turn more to pork or chicken or even go meatless.

On using game theory to better understand and resolve DevOps issues (with help from LLMs)

I have seen a consistent set of issues when it comes to DevOps implementations.  The issues are:

  • Clients say they are emploing a DevOps lifecycle, but it’s more the case they are simply doing continuous integration and deployment (CI-CD) supported by an associated tool chain.
  • Many times the Dev (development) teams are separate from the Test teams and the Ops (operations/support) teams. Sometimes they are even in different silos in the organization. As a result, communication and collaboration — an critical part of a successful DevOps implementation — may be limited and weak.
  • As a result of that separation, the incentives may differ by team. In some cases they can end up almost competing against each other. (For example, I have seen Dev teams take time away from the Test teams in order to complete their assigned development tasks. The Devs “win” and the Testers “lose”.)
  • Quality standards held by the Test and Ops teams may not shift left to the Dev teams. As a result, quality responsibilities may lie mostly with Test/Ops, or there is a significant imbalance. (The Devs may do some limited unit or regression testing, but not enough to prevent bugs from being passed downstream in the DevOps lifecycle.)
  • Delivered functionality on time may become the primary driver for a project in order to meet the needs of key stakeholders, while the non-functional / quality of the system (e.g., performance, observability) may become secondary or even considered a Day Two problem.
  • As a result of all this, it’s not surprising to see Dev teams become focused and measured on delivering functionality at increasing velocities, while the Testers and to some degree the Operators have to maintain that focus / those measurements, while also being responsible for and measured on software quality.
  • In the end, you can get a situation where Devs are measured in terms of speed and functionality, while the Testers / Ops are measured on speed, functionality plus software quality. If delivering software is like a game, the Testers and the Operators are playing the game with a handicap due to quality metrics.

I have thought about these issues for some time and I wanted a framework to apply to them so I communicate in a more concrete way not only the issues but how to address them. I thought that one such framework to use could be supplied by game theory.

To apply game theory to DevOps, I started by considering the Devs as one set of players and Testers/Ops as another. To quickly work through some scenarios to see if such an approach was even valid, I decided to use LLM models. First I used GPT-OSS-20B to get my initial results. While it seemed promising, I noticed some errors creep into its analysis with regards to Nash equilibriums. I swapped out that model and used a larger LLM (provided by IBM Bob) to analyze and correct the initial results and produce a final analysis. You can read that final analysis, here.

I learned a few things from this exercise:

  • when the Devs are not bound by quality, a Nash Equilibrium is achieved, but it is one where they get a bigger payoff for going as fast as possible and the Testers/Ops have to settle for a situation where they suffer and get a worse payoff.
  • If the Devs are bound by quality, they get a lesser payoff but the Testers/Ops  get a better payoff and overall software quality can be increased. It seems obvious to do this, but working out a pay-off matrix helps with this.
  • Under certain conditions the Nash equilibrium can become weak for the Devs and that may result in them being less quality focused in order to succeed. Again, not surprising.
  • Game theory can show why Developers rationally can decide to forgo quality, but it may not be the best way to practically analyze your project. I’d simply recommend you continually work to spread quality measures throughout your DevOps lifecycle and to continually work to improve quality as the software matures.
  • LLMs can speed up the production of documentation just like it can speed up the production of code, but you don’t want to solely depend on it to get the documentation correct. Based on my understanding of game theory, I think its output is correct. If someone with a better understanding wants to correct me, I’d be happy to hear about it and I’ll be happy to amend this post.
  • For a more important project I would heavily modify the LLM documentation and take more ownership of it. This is more of a side project / exercise for me, so I decided to leave it as it is and move on to bigger things with bigger payoffs, to use the language of game theory. I hope you can still find it useful. I’d suspect the number of people who read it can be measured in one hand.

(P.S. The opinions expressed in this post are mine and do not necessarily reflect those of my employer.)

On the decline and fall OpenAI?

Is this the beginning of the decline and fall of OpenAI and their leader Sam Altman? I thought that recently, when I read how Apple was suing OpenAI for supposedly stealing its secrets.  And that’s not the only lawsuit against them: the New York Times and other media companies are also suing them for theft. (Elon Musk also sued OpenAI  but lost. A good run down on that lawsuit is here.)

Now I am thinking differently. I wanted to say the decline started with staff breaking away and forming Anthropic in 2021, though even that isn’t straightforward.  This loss of former staff has been mitigated by other things, like large investments from Microsoft. Those investments, the growing user base and the lack of scrutiny from the press gave the impression that all was well at the powerful company. (As recently as 2023 publications like the Times and the Wall St Journal were providing highly positive coverage about Sam Altman and his company and he and OpenAI were the darlings of the tech world.)

I am still inclined to think that competition from Anthropic which will lead to losses for OpenAI. It’s still difficult to tell, but observers like Motley Fool are saying Anthropic is pulling into the lead and a gap is opening. I believe that gap is only going to widen.

In addition to Anthropic, there is other competition coming from China and elsewhere. That increasing competition and the declining Microsoft support will also put a squeeze on OpenAI. (You can see from this piece, which notes that Microsoft is moving towards their own models, while others are looking to Chinese models for business use.)

Losses or not, perhaps OpenAI will not suffer a decline but a displacement from the top spot. Perhaps OpenAI can become profitable and grow despite having to share the AI marketplace with others. (It works for Apple: they own 1/5 of the smartphone market and they are doing fine).

Beside competition, another thing that could cause their decline is the leadership of Sam Altman. No one is providing positive coverage about him now, as you can see from this and this and this and this and more. Even if all that is true, it would not be the first time a terrible leader has lead a successful IT company. So perhaps OpenAI can succeed despite being lead by the current CEO.

I started writing this piece thinking: OpenAI is going to fall, and I wanted to put a marker in this blog to note that. By the end, I am undecided. I’ll put this here for now and revisit things in a year.

P.S. Thanks for reading this. I wrote this so I could learn what I thought. Even though I changed my mind in the writing, I still think it is worth putting down this marker for now.

On parenting, getting older, torture, and being useless (i.e., interesting things I wanted to write about and maybe will some day, 2026 edition)

Every year I like to let go of these links I wanted to write about but never found a way to and likely never will. They’re interesting still. Last year it was 60 or so. This year it’s a little less.

A few travel stories I liked were this, 21 US airports that let you walk to a gate without a ticket, and this, The longest train journey in the EU.

I enjoyed several historical pieces, such as this 13th-Century Cookbook Featuring 475 Recipes from Moorish Spain that recently was Published in a New Translated Edition, and this Historical Photos Taken Just Before The Iconic Moment. I found this piece from February 1970 in the Vassar Newspaper & Magazine that was critical of IBM during the Vietnam war to be worth a read.

Lloyd Alter shares interesting ideas on his Substack. For example, this one on how it’s time to dump Railway Time and War Time and go Local Time, and this one on why we need deposits on everything.

Cincerning matters of privacy, I thought this collection of Dark Patterns and Unethical Designpatterns were Ring Cameras, with the Ring company dark patterns is this: ICE, Secret Service, Navy All Had Access to Flock’s Nationwide Network of Cameras and this Zenni’s Anti-Facial Recognition Glasses are Eyewear for Our Paranoid Age. No doubt such abuses lead people to ask Hackaday: How Do You Detect Hidden Cameras?

I started to collect articles on getting older, like this one, When You’re ‘Fired From Retirement’. It was a bit grim. Less grim and easier to read is this post about a 72 year old woman who graduated medical school and was starting  residency.  This was also positive: I’d Like to Stay 85 Forever.

Parenting is a difficult job, to say the least. Some might say impossible at times. I thought that when I read this, consider the parents caring for adult children not in work or education, and also this, This could happen to us: Parents who see themselves in the Reiner case

I don’t imagine too many will want to read about Lee Geun-an, Infamous ‘Torture Master’ Under South Korean Dictator who recently died at 88. I like to remember things like that. Likewise, I like to remember what Michael Ignatieff said about torture in Macleans: here. Never forget.

Finally,

I liked this, on Nudge theory. Also…

 

On the Cagots


By accident I came across this article in Jstor on the Cagots, the forgotten untouchables of France. They suffered discrimination  for many centuries, but no one knows seems to know why, although there are many theories. They were erased as a group from France through assimilation, even though up until the beginning of the 20th century it was shameful for a non-Cagot person to marry a Cagot.  You can still see evidence of the discrimination they felt in churches in France, where the Cagots had a separate entrance and separate fonts.

All in all, a fascinating story of a mysterious people. I find them especially fascinating because this shows how a group can suffer discrimination and persecution, even though there is nothing different about them, other than the fact they can be labelled and considered worthy of such poor treatment.

For more on the Cagots, see wikipedia. For a story on someone who claimed Cagot ancestry, see this.

Friday Night Cocktail: the Lower-Alcohol Dirty Martini

I have not published a Friday Night cocktail post in some time. Possibly because I have been drinking much fewer cocktails than I did a few years ago, especially during the pandemic. While I have enjoyed many different cocktails in that time, my preference tends to be a martini, which I have often posted about here.

Ley me add one more post of a recipe for a particular form of martini because it captures the zeitgeist: it has a low A.B.V. and it is not just A martini but A DIRTY martini. If you are like many others, those two things will appeal to you. To be honest, I am curious about this myself!  Now I still wouldn’t start making them on a school night, but YMMV.

Whenever you have one, I hope you enjoy it. Cheers!

On the problem of popular restaurants not taking reservations


I get why popular restaurants don’t want to take reservations. Case in point, the very popular Chubby Fish restaurant in Charleston, S.C. According to this piece:

Chubby Fish doesn’t accept reservations because London (James London, Chubby Fish chef and co-owner) doesn’t want the restaurant to be exclusive. The goal is for people to be able to show up at 4 p.m. and be able to eat dinner the same night.

In some way that makes sense: you don’t have empty tables ever and the people who really want to go will line up.

Alas there is a problem:

Now that some people are paying extra for line waiters to arrive early and claim desirable dining times, others in line aren’t getting the same opportunity they once had. As a result, they’re having to arrive earlier and earlier, London said.

“We never wanted that for our guests. It’s not the hospitality we wanted to display,” he said.

Recently, London spoke with the line-waiting businesses…about limiting the number of reservations they’re allowed to procure at a time, he said. But he worries about what could happen if the businesses continue to grow.

“I don’t want to be forced into having to take reservations to eliminate the professional line-waiting services, but what does it look like three years down the road?” he said.

And how much are these line-waiting services charging?

Ella Ward, a senior at the College of Charleston, realized that made for a good business opportunity….Ward has turned waiting in line at Chubby Fish into a thriving side hustle. She charges $60 on weekdays and $80 on weekends to wait out the line and snag seating times on behalf of others. She recruited other CofC students to help her manage requests, and the team is booked through June.

This is just a form of line jumping and is not unlike someone slipping FOH some money to get a good table. And not for a small amount either. If everyone is paying $60 or more to get a seat in your restaurant, you are an exclusive restaurant, whether you like it or not.

No reservations (i.e., supply) is fine for restaurants that have roughly the same demand. But if demand vastly outstrips supply, you are going to have problems like line-waiting services. You are going to be transformed into an exclusive restaurant by the marketplace. In the era of social media and vast tourism where the marketplace is gigantic, this is going to become more of a problem for some restaurants like Chubby Fish.

 

Pressing pause on the newsletter (i.e. what would have been the June 2026 edition of my not-a-newsletter newsletter)

The first half of 2026 is done. I was planning to write a recap over what has been happening since December, when I posted my last newsletter. Today I decided to change that plan.

I’ve been posting newsletters regularly since my first one in March 2020. Back then, they were monthly: lately they’ve become biannual. Back then, they gave me something to do in lockdown and they gave me a way to track the pandemic. Lately they have turned to the Trump administration, since like COVID-19, it has a big impact on the world.

Unlike the pandemic, however, I don’t feel like tracking or writing about Trump and his administration any more. Perhaps I will at year end, after the midterms, when hopefully elections do to Trump what vaccines did for COVID-19. Perhaps things will feel different then and I will feel different and want to write about what he has done to America and the world. Until then, I think I’ll focus on other things on this blog.

I’m glad I posted about the pandemic. I’m glad I stuck with posting my newsletter here on this blog and not on a separate Substack newsletter. I find it interesting to go over back over them from time to time to see how things have changed. For anyone interested, you can find them all here.

It’s good to persist in an activity. It’s good to know when to change direction too.

 

 

Things are tough? You need Halls as a coach!

I realized recently that Halls cough drops do more than assist with a cold. They also assist with your mental health! These are some of the phrases I’ve captured on the wrappers than come with each lozenge:

  • You’ve survived tougher
  • Nothing you can’t handle
  • Go for it
  • Be unstoppable
  • Get back in there champ!
  • You can do it and you know it
  • Don’t try harder. Do harder!
  • Get through it
  • Don’t wait to get started

Lots of good encouragement there. And the exhortations are in English and French. Formidable! 🙂

Kidding aside, I think this form of message is not bad. Any little thing to help you feel less miserable is a good little thing.

How to win the New York Times Crossplay game against the computer in hard mode (June 2026)

I have been playing the New York Times Crossplay game against the computer in hard mode and most of the times I win. Here’s what I’ve discovered:

  • In hard mode the computer “likes” to play longer words whenever possible. Many times these are uncommon words.
  • The computer will try and play the 2W and 3W tiles, but it seems to “prefer” to play longer words over shorter words, even if a shorter word on a 2W or 3W might gives more points.
  • The computer will stack shorter words but it tends to do that later in the game when it is harder to play longer words. It will stack longer words if it can.

These have been my observations. Give that, to increase my chances of beating the computer, I will:

  • hold back letters with higher values unless I can score at least 20 with them on my current turn.
  • wait for the computer to set me up for 2W and 3W tiles and then play the higher value tiles. I find it will play a long word to maximize it’s length, even if it is easy for me to stack on top of its word and get the 2W or 3W tile. It doesn’t seem to play defensively.
  • My expectation is the computer will do this at least 2 or 3 times a game, so being patient even when I am down is key to beating the computer.
  • if I can’t do anything good with an open 2W or 3W, I will take it if there is a chance the computer can reach it in 5 tiles.
  • I will set myself up for a 2W or 3W in the next move by holding back a tile. Let’s say I can end a word in S, but it means it is possible for the computer to use that S to play a word and to get to a 2W or 3W tile. If I was playing a person, I might be worried they were going to get that in the next round. But if there are open spots on the board where the computer can play a longer word, chance are it will go for the longer word instead, so in the next turn I can lay down the S and hopefully spell out a word to get the 2W or 3W tile.

I say most of the time I will beat the computer in hard mode. Crossplay is a game of chance as much as anything, and some games I will get a bad assortment of letters and the computer will get a great assortment of letters and I will lose. But by adopting the approach I described above, I can frequently beat the computer in hard mode, even though it knows more words than I do.

P.S. My overall general approach is to always aim for the highest score for each play. Often this means stacking words — even small words like “LI” and “OR” — on top of other words to maximize my score rather than playing long words, the way the computer prefers to do. Stacking words gives less openings for my opponent, which is especially effective against the computer, which will often open up the board by playing a long word.

How not to clean your house

While I have written a fair bit about cleaning on this blog — a chore I do not like — today I want to share something different, how not to clean. Specifically, 7 TikTok cleaning hacks to avoid according to The Washington Post. Some of the hacks are a waste of time, others could potentially kill you.

The hacks to avoid are:

  1. mixing cleaning products together – potentially lethal
  2. Mopping floors with laundry detergent
  3. Cleaning upholstery with a detergent pod and boiling water
  4. Boiling Pine-Sol – also potentially dangerous
  5. Vacuuming with laundry scent beads
  6. Putting laundry products such as detergent pods and scent beads into decorative jars for aesthetic reasons.
  7. Fabuloso in the toilet tank

For more detail, head over to the Washington Post.

On nominal amounts

I became interested in the topic of peppercorn rent when I read how the English Royals use it. That lead me to reading these two pieces (Peppercorn Rent: A Comprehensive Overview and What Is Peppercorn Rent – and why is it in the news?) I recommend them both.

That topic lead me to think of other such forms of nominal fees or payments. A good example of which is the One-dollar salary. Very nominal indeed.

 

A small note on Musk and SpaceX and the Mongolian Horde Approach

Just wanted to add a small note on SpaceX and the insanity around it all. Paul Krugman and others have been dismissive of the value of SpaceX. Is it a meme stock? Is it a scam? I can’t say.

I can say this. The value of SpaceX allows Musk to deploy his Mongolian Horde approach (something I wrote about here). In short, Musk solves problems by throwing massive resources at a problem until he achieves his objectives. So if he wants xAI to be successful, he can buy Cursor or whatever else he needs for it to be successful. If that doesn’t work, he can keep burning through money and acquire what he needs to get to where he wants to go.

I can’t say he is going to be able to do this indefinitely, but I think he can do it a lot longer than people thing he can. If indeed he is a Ponzi scheme, I do not believe he is one in danger of collapsing any time soon. But let’s see!

On Ponzi Schemes

Charles Ponzi

Over on his Substack, Paul Krugman makes the case for Elon Musk being a Ponzi Scheme. He says: “We have a term for enterprises that look successful because they keep drawing in new investors and keep drawing in new investors because they look successful. They’re called Ponzi schemes. And Elon Musk is basically a human Ponzi scheme.”

I don’t know. Musk is many bad things, but I think he is more than a Ponzi scheme, based on this entry in Wikipedia. Now Bernie Madoff: his enterprise was a Ponzi scheme. But I think there is still some substance behind the enterprises that Musk is running.

Paul Krugman obviously know economics, but I think he is stretching things to insult Musk. Check out these linked pieces and decide for yourself.

P.S. Elon Musk and his companies may not be a Ponzi scheme, but this piece in Techdirt makes the case for how they are a scam, in particular SpaceX.

(Image of Charles Ponzi from Wikipedia)

Bernie’s rule of business predictions of 5 years (or more)

If someone makes the following business prediction: event X will happen in 5 (or more years) think to yourself: they absolutely have no idea.

 

As I said in an earlier post:

For any business prediction, the time frame determines the probability. If they say in a year, they either are doing it themselves or someone they know is. If it is 1-2 years, it won’t happen this year, but someone is working on it. If they say 3-5 years, then it’s likely not in progress, but there is talk of working on it.

The only prediction that is useless is 5-10 years. If someone says 5-10 years, they are saying something like “I don’t want to say it will NEVER happen, but it is not even close to coming to fruition, so I will predict 5-10 because who is even going to come back and ask me about it in a decade from now?” 🙂

Books for reading when you’re going through hard times

The New York Times asked experts to “share the literature that has helped them cope with death, illness and despair” and the result was this article: 5 Books to Lift You Up During Life’s Hardest Times (gift link). You may have these or have heard of them already:

  1. ‘Being Mortal,’ by Atul Gawande
  2. ‘Meaty,’ by Samantha Irby
  3. ‘Gilead,’ by Marilynne Robinson
  4. ‘Blindsided,’ by Richard M. Cohen
  5. ‘Vigil,’ by George Saunders

If not, check out that gift link and find out more about the books, including why and how they can lift you up.

What’s better than 150 Extra Engineers? This blog post! (What I find interesting in tech, June, 2026 edition)

Ok, technically the IBM 604 Electronic Calculating Punch was better than 150 extra engineers. I can’t provide you with one of those, but I can provide you with this list of links on what I find interesting in tech lately (besides AI, which I have posted on separately since it is a big topic in itself).

Hardware

Software:

Raspberry Pi :

Miscellany:

(Credit for the top image: File:IBM 150 Extra Engineers 1951.jpg – Wikimedia Commons)

AI: what people are doing and what people are saying (June, 2026)


When it comes to AI, there are 1) lots of people doing practical things with AI and 2) lots of people writing about AI. They rarely overlap. 🙂 

But overlap they do in this post! For those of you not doing practical things with AI, skip to the second section.

1) Here’s a bunch of links for people who are doing practical things with AI:

2) Here’s a bunch of links for people who like to read about AI:

How to be an artist, how to do art, and what art can do for you

 

As a frustrated artist, I not only gather links about artists, but I also gather links about art. Here are just some of the ones I’ve picked up recently:

If you want to be an artist, it might be good to ask established artists what their thoughts are. Artsy.net did a series on this, and I put together a list of just some of them:

Also helpful, swissmiss has a good piece to help you determine the type of artist you are / want to be:Artist vs Creator

If you decide you want to be a painter, here’s some things to get you started. Sometimes all you need to do is to start. (These are taken from the book Learn to Paint in Acrylics with 50 More Small Paintings: Pick Up the Skills, Put on the Paint, Hang Up Your Art. I recommend it and Learn to Paint in Acrylics with 50 Small Paintings: Pick up the skills * Put on the paint * Hang up your art):

Maybe collage is your thing. If so, check out:

I am a big fan of trois crayons drawings, so I liked this: Chalk it up to three colors: bring together color, form and design by using the trois-crayons technique–traditionally only red, black and white chalk

I’m also a fan of zines, so I thought this was good: Gen Zine: DIY publications find new life as a form of resistance against Trump | US news | The Guardian

Conceptual art isn’t for everyone, though it is for me. Perhaps you might want to consider that route: Conceptual Art 101: A Beginner’s Guide | DailyArt Magazine – www.dailyartmagazine.com

Finally, kintsugi is more than just a way of repairing things. To see what I mean, check this out: Trevor Noah Explains How Kintsugi, the Japanese Art of Repairing Pottery, Helped Him Overcome Life’s Tragedies | Open Culture – www.openculture.com

From the todo lists of Haring to the cutouts of Matisse as well as some stairwell drawings by Byrne (the arts on Wednesday, June 2026 edition)

Here’s the latest edition of #theartsonwednesday, a mini-newsletter within my blog highlighting recent stories on visual artists new and old that I’ve found interesting:

 

On the ubiquity of Paul McCartney

Paul McCartney is everywhere, it seems. He played 2 big shows in LA in March. He was at the recent closing of The Late Show with Steven Colbert. Not only was he a big part of the 50th anniversary celebration of SNL, but he closed out the 51st season. Now he has a new album coming out, “The Boys of Dungeon Lane”, so he’s on every platform promoting that, including The New York Times and The Guardian.

When he’s not performing current work, he’s plugging away at the past. We recently saw the release of “Man on the Run”, a documentary of his days in his band, Wings. (Trailer here, reviews here by the Guardian and the New York Times). And then there is his photo exhibit, “Paul McCartney Photographs 1963–64: Eyes of the Storm” at the AGO.

He’s in his 80s. He’s not the singer he was in the 60s — see points 13 and 14 here to see what I mean — but he still puts on a good show. He still looks healthy and fit. Hopefully he will be with us for some time to come. Lord knows he is not lacking for opportunities to perform.

P.S. For more on McCartney in Wings, check out this (Wings by Paul McCartney review – a brilliant story of post-Beatles revival | Paul McCartney | The Guardian)  and  this (Paul McCartney: where to start in his solo back catalogue | Music | The Guardian).
Here are some odd and ends in terms of links I found which I am including in this post, to stick in my back pocket so to speak: ‘The Beatles Anthology’ Gets a New Ending: Director Oliver Murray on How Episode 9 Finally Put a Sweeter Cap on the Group’s Historic Docuseries, plus was Denny Laine ever properly paid (I think so) and also I thought this piece on his 3 hour show in Hamilton last year was good.
Lastly, I’ve written quite a bit on Paul McCartney. You can find it, here.

Back to Dirac (What I find interesting in math and science, May 2026)

 

I am not doing as much study in the area of math and science lately and haven’t written on the topic since last May, but I still found time to note these pieces and wanted to share them. A typical smattering of astronomy, physics, math, biology, what have you….

 

 

On developing a habit of writing politicians

Over the years I’ve developed a habit of writing politicians. I shared this habit with my friend Norbert and I after that chat I decided to share it here in case you want to develop a similar habit.

If you want to write politicians like me, you need to know how to contact those pols that mean the most to you. For me those three are the mayor, the premier of the province and the prime minister. The mayor is reachable by email so I have that address handy. The other two are reachable via a web form, so I have that info in hand also.

I have a template that I use for letters I write to each of them. When I want to write them about an issue, I pull up the template and fill it in. I usually say what the issue is, then what I would like them to do about the issue. Then I send it.

That’s it! It’s a two step process. Very quick. Occasionally I will follow up, but the purpose is really to register my voice. I don’t expect anything other than a canned response. Only one politician’s team, Mayor John Tory, ever made an effort to respond in a more personal manner.

While I do write about issues, I will also write politicians of all sorts and compliment them if they do something I approve of. Typically these will be constructive actions and often ones where politicians are working together.

I write politicians in a style I use for anyone who works for me, because they do. I don’t insult them or get angry with them: I find that never works when it comes to influencing someone I want to change.

Ideally I would write at least one politician a week but it’s more the case I write one a month.

I hope you can take up this habit. It makes me feel like I am doing something small but important in our democracy. Just like voting is small but also important. Good luck with it.

What are the best novels of all time? What about the best Shakespeare plays?

What are the best novels of all time? What about the best Shakespeare plays? And what do these two lists have in common?

What they have in common is that The Guardian has taken a crack at both of them:

This is smart in a few ways. One, there is nothing that gets people going than someone saying: these are the best. No doubt there will be many people checking them out (and driving traffic to the Guardian web site) just to see if they agree or disagree. Two, it’s really good thoughtful that the contributors have made to come up with such lists.

I grew up with The Book of Lists, so I am always interested in a good list. If you’re similar, or simply a fan of the Bard of a great novel, then check them out.

Restaurants loved and lost: Libertine in New York City, Rodney Scott’s in Charleston

The minute I read this review by Pete Wells, I knew I had to go Libertine. I love a French bistro generally, and this one sounded especially good. On the night we went, everything about it was ideal. We snagged a lovely table near a window and had a waiter that bent over backwards to make sure we had a memorable evening. The food, from the appetizer (oeuf mayonnaise) to the mains (that sausage and potato dish) to the a dessert (a massive chocolate mousse), was delicious. As soon as we were home for the evening, I was already dreaming about when I could go again.

Sadly those dreams are done, for Libertine is closing this month. The same people own the space and plan to turn it into something else which will no doubt be great. But I would have liked another chance to eat at that fine New York City bistro.

Another place I had hoped to return to some day was Rodney Scott’s in Charleston. It too was a place that I heard of through someone famous (Anthony Bourdain) and I had a chance to dine there once at it’s location on King Street. I filled a plate with lots of food that Scott was known for, from the BBQ meats to the tasty sides. Alas there are almost too many great places to dine in Charleston and I never did go again, and now I never will.

I look forward to finding out what goes next in the space where Libertine was. And I hope Rodney Scott can make a comeback.

P.S. For more on Libertine closing, see this. For more on Rodney Scott’s closing, see this.

P.S.S. June 19, 2026: looks like Libertine is being reborn as Capitaine. Look forward to going some night!

How to set up a blog that’s just a simple html page

I wanted to go back to the very beginning of blogging, where someone’s blog could consist of just a simple html page with maybe some images to go along with it. But I also wanted to have an easy way to update it. I’ve succeeded in doing just that, with the result visible here: https://berniemichalik.com/simpleblog.html

If you think this is something you want to do, head over to here and check out what I say in that repository. It’s a pretty basic setup. There’s even advice how to host it in a bucket on Amazon, which is what I do.

Let me know if you have any feedback, especially if you think things aren’t clear. Thanks!

On AI generated code and the DevOps Lifecycle

 

It’s an exciting time to be a software developer.  Exciting and weird.

It’s exciting because of what developers are capable of doing with agentic AI tools from companies like Anthropic. It’s weird because as Clive Thompson writes: “In the era of A.I. agents, many Silicon Valley programmers are now barely programming”. Yet while programmers may be barely programming, they are still producing code.

Indeed, not only are they producing code, they are producing A LOT of it. Take this example in this other piece by Mike Isaac and Erin Griffith:

When a financial services company recently began using Cursor, an artificial intelligence technology that writes computer code, the difference that it made was immediate.

The company went from producing 25,000 lines of code a month to 250,000 lines. That created a backlog of one million lines of code that needed to be reviewed, said Joni Klippert, a co-founder and the chief executive of StackHawk, a security start-up that was working with the financial services firm.

And it’s not limited to senior programmers. As Isaac and Griffith write:

“The blessing and the curse is that now everyone inside your company becomes a coder,” said Michele Catasta, the president and head of A.I. at Replit, an A.I. coding start-up in Foster City, Calif.”

It all sounds great: who wouldn’t want such a productivity gain? Alas, there’s a problem. Isaac and Griffith add:

““The sheer amount of code being delivered, and the increase in vulnerabilities, is something they can’t keep up with,” she said. And as software development moved faster, that forced sales, marketing, customer support and other departments to pick up the pace, Ms. Klippert added, creating “a lot of stress.”

A lot of code means a lot of stress. Not just for sales, marketing and customer support teams, either. I suspect it’s also causing plenty of stress for testing, operations and application support teams in organization that rely on a DevOps lifecycle for software development.

To deal with such stress, any organization that adopts AI technology to write code and also uses DevOps needs to consider their entire software development lifecycle in light of such code.  AI generated code, no matter how fast it is coded, will have limited value if it’s not reviewed and tested properly. It will have limited value if it is not written in a way that it can be supported by applications and operations. Not to mention it will be hard for others to fix and enhance it later if it is not built and deployed using agreed upon processes by the organization.

To avoid creating unsupportable code with problems that will accelerate technical debt, developers of all levels need to know more about their code than “it works on their machine”. If such developers expect others within their organization to use their code, they must write code that doesn’t impact their organization’s DevOps lifecycle. They must take the time to write AI generated code that is supportable. That additional effort by people generating the new AI code is essential if organizations are going to survive the code tsunami that is coming.

The good news, I think, is that everyone that participates in the DevOps lifecycle will be able to devise ways to support AI developers such that their code will fit in nicely.  After all, as this piece by IBM on the DevOps lifecycle explains: “The DevOps process describes how software moves from ideation, through production and feedback, and back to ideation, with development and operations teams working as a single, collaborative unit.” Organizations need the new AI supported developers working with others as a single collaborative unit to succeed. If they do, the new productivity will go from something that causes stress to something that causes success.

(The opinions expressed above are mine only and not necessarily those of my employer.)

Microchanges: or how to get unstuck when you are stuck with the 10 minutes a day principle

I was feeling stuck lately. I thought I needed some goal or purpose to get me going again. But what?

Lately when I’ve been stuck with personal challenges I’ve been turning to my own personal AI (i.e., Ollama running the LLM gpt-oss-120b), for advice. I asked the software: What are 3-5 goals I could set for the month that other people have found worthwhile and that could help me gain momentum to set and achieve other goals?

I got back the following:

Below are five monthly goal ideas that many people swear by because they’re concrete, doable in a single month, and give you a quick win that fuels the next big project. Pick one (or a mix) that feels most relevant to you and turn each into a SMART target (Specific, Measurable, Achievable, Relevant, Time‑bound).

# Goal Why It Works How to Structure It
1 Daily “micro‑habit” – e.g., 10‑minute meditation, 5‑minute journaling, or 30‑sec body‑weight routine Small, consistent action builds a rhythm. Once the habit feels automatic, it’s easier to layer more complex goals on top. Specific: “Meditate at 7:00 a.m. every weekday.”
Measurable: Track days on a habit‑tracker app.
Reward: After 21 consecutive days, treat yourself to something you enjoy.
2 30‑Day “Read a Page a Day” challenge Reading a page a day adds up to ~30 pages a month – enough to finish a short book or digest a chapter of a non‑fiction title. It keeps your mind curious and gives you a tangible end‑product. Specific: “Read 20 pages of Book X each day.”
Accountable: Post your page count on a public calendar or share a weekly summary on social media.
3 Declutter One Area per Week Cleaning a single spot (desk, closet, email inbox) gives an instant sense of order and control, motivating you to tackle the next spot. Specific: “Sort the top shelf of my bookshelf every Sunday.”
Measurable: Log each area with before/after photos.
Result: After 4 weeks you’ll have a noticeably tidier environment.
4 Save/Track a Small Amount of Money A financial win, even a modest one, builds a habit of budgeting and shows that you’re moving toward a bigger financial goal. Specific: “Set aside $25 each week into a savings account.”
Track: Use a budgeting app to see the cumulative total at month‑end.
Celebrate: Use the savings for a fun treat or a future investment.
5 One New Skill Minute a Day Dedicating just 15 minutes daily to a new skill (language, guitar, coding) creates incremental progress that feels fast and keeps you engaged. Specific: “Learn 5 new Spanish words every morning.”
Measure: Keep a vocabulary list; test yourself weekly.
Progress: At month‑end you’ll have a solid foundation to expand.

Quick Tips to Keep Momentum

  1. Track it visually – a wall calendar, habit‑tracking app, or a simple check‑box list. Seeing your streak grow is a powerful motivator.
  2. Start with the first week – often the hardest part is just getting started. Commit to “just 1 minute” or “just one page” to overcome the inertia.
  3. Set micro‑rewards – a favorite coffee after week 2, a new playlist after week 4, etc. Rewards reinforce the positive loop.
  4. Reflect weekly – spend 5 minutes reviewing what worked and what didn’t. Adjust the goal slightly if it’s too easy or too hard.
  5. Share your goal – tell a friend, post on social media, or join a community. External accountability can keep you honest.

Pick a goal that feels both meaningful and achievable for this month. Once you finish it, you’ll have a proven “win” to celebrate, and that confidence will make launching your next big goal a lot easier. Happy goal‑setting!

Wow. That was an actual helpful list. I’ve adopted all five just to see how I can do.

  1. Microhabit: 5 minutes of kettlebell exercises and squats
  2. 10 minutes of reading a day
  3. 10 minutes of declutter a day (vs a week)
  4. 5-10 minutes of financial tracking
  5. 10 minutes of drawing / art practice

What I’m noticing at first:

  • The short time periods allocated for each one makes it hard for me to skip it, so I haven’t been, even on days I am tired or busy.
  • Even short periods of activity add up. I can do much more exercise than I could at first. My reading is making progress.
  • The friction that prevented me from doing things have decreased.
  • I use a physical log book to keep track. This helps reinforce the sense of making progress.
  • It’s better to do things poorly to keep the momentum then to get hung up on poor results for a given day.
  • I was getting bored going into the third week.
  • Cumulative results were good.
  • Switch things up to keep the momentum going.

What I observed towards the end:

  • I was getting stronger but levelling off.
  • Reading is easy to be consistent and if anything I tend to have periods where I read longer (e.g. if the book I am reading is engaging)
  • Some things make me want to go deeper: example on finances.
  • The drawing is more scattered. It is more creative, less rote, unlike the weights or even reading.
  • I wanted to expand this to 6 or 7 microhabits a day but I was running out of time.
  • It was easy to stick to the microhabits when my schedule was light: it got harder when I got busy. No doubt that would be less so if I took on 2 or 3 instead of 5.
  • Doing multiple things made me stick to things more than doing one thing would have.
  • Eventually I want to making things a project. For example, on my decluttering of my room, that took a long time. A complex mix of things there. In places like the kitchen and other spaces, less sentimental items, so easier to declutter.

I stopped doing this once I took a break and went on vacation. It was rewarding, though. And it helped jump start me into doing things I was not. I highly recommend it.

P.S. For more on the importance of doing things 10 minutes a day, read this.

(Image from Valeriia Miller)

If you had to buy only one ebook…

If you had to buy only one ebook, I would recommend this one: 50 Masterpieces You Have to Read Before You Die. Coming in at just over forty thousand pages, it has the following:

  1. Alcott, Louisa May: Little Women
  2. Austen, Jane: Pride and Prejudice
  3. Austen, Jane: Emma
  4. Balzac, Honoré de: Father Goriot
  5. Barbusse, Henri: The Inferno
  6. Brontë, Anne: The Tenant of Wildfell Hall
  7. Brontë, Charlotte: Jane Eyre
  8. Brontë, Emily: Wuthering Heights
  9. Burroughs, Edgar Rice: Tarzan of the Apes
  10. Butler, Samuel: The Way of All Flesh
  11. Carroll, Lewis: Alice’s Adventures in Wonderland
  12. Cather, Willa: My Ántonia
  13. Cervantes, Miguel de: Don Quixote
  14. Chopin, Kate: The Awakening
  15. Cleland, John: Fanny Hill
  16. Collins, Wilkie: The Moonstone
  17. Conrad, Joseph: Heart of Darkness
  18. Conrad, Joseph: Nostromo
  19. Cooper, James Fenimore: The Last of the Mohicans
  20. Crane, Stephen: The Red Badge of Courage
  21. Cummings, E. E.: The Enormous Room
  22. Defoe, Daniel: Robinson Crusoe
  23. Defoe, Daniel: Moll Flanders
  24. Dickens, Charles: Bleak House
  25. Dickens, Charles: Great Expectations
  26. Dostoyevsky, Fyodor: Crime and Punishment
  27. Dostoyevsky, Fyodor: The Idiot
  28. Doyle, Arthur Conan: The Hound of the Baskervilles
  29. Dreiser, Theodore: Sister Carrie
  30. Dumas, Alexandre: The Three Musketeers
  31. Dumas, Alexandre: The Count of Monte Cristo
  32. Eliot, George: Middlemarch
  33. Fielding, Henry: Tom Jones
  34. Flaubert, Gustave: Madame Bovary
  35. Flaubert, Gustave: Sentimental Education
  36. Ford, Ford Madox: The Good Soldier
  37. Forster, E. M.: A Room With a View
  38. Forster, E. M.: Howards End
  39. Gaskell, Elizabeth: North and South
  40. Goethe, Johann Wolfgang von: The Sorrows of Young Werther
  41. Gogol, Nikolai: Dead Souls
  42. Gorky, Maxim: The Mother
  43. Haggard, H. Rider: King Solomon’s Mines
  44. Hardy, Thomas: Tess of the D’Urbervilles
  45. Hawthorne, Nathaniel: The Scarlet Letter
  46. Homer: The Odyssey
  47. Hugo, Victor: The Hunchback of Notre Dame
  48. Hugo, Victor: Les Misérables
  49. Huxley, Aldous: Crome Yellow
  50. James, Henry: The Portrait of a Lady

And it’s not just for Apple ebook readers: there’s a Kindle version too.

Best of all, it costs less than three bucks!

A few thoughts on Palantir as dangerous uncool jokers

Barbara Kruger once slammed some as “a ridiculous clusterfuck of uncool jokers”. I think of that phrase often, and I thought to apply it again when I read about Palantir’s manifesto on a recent thread on X. It’s easy to dismiss it as “the ramblings of a supervillian”, but I think there is something more to it. Reading it, I thought I saw  references to ideas like those espoused by Nazi thinker Carl Schmitt, among others. I mean, Palantir wouldn’t be the first on the American right to do that.

As an experiment, I asked ChatGPT what German writings from the first half of the 20th century this resembles. It came back with this:

 

That was a good summary of the manifesto. Reading more on Spengler, Jünger and of course Schmitt confirmed that. The Palantir manifesto might not be a Nazi document, but it is aligned with writings of pre-1933 radical conservative thought, much of that which went on to influence Nazi Germany. To me the manifesto shows how dangerous Palantir is at its core.

Now I get it: for some of you, it’s hard to take these jokers at Palantir seriously. I mean:

 

Not to mention things like the fact that they are taking the time to design their own…chore coat, or that Palantir associate Peter Thiel is saying such crazy things that the Catholic church is mulling burning him at the stake.

So yeah, while they may be a bunch of uncool jokers, as far as jokers go, they are far from harmless. Being a joker and being dangerous aren’t mutually exclusive.

 

On thinking about the U.S. national debt exceeding 100% of the GDP (gross domestic product). Again.

Every so often the debt of a country becomes a political issue and writers start cranking out pieces like this in the Wall Street Journal about how:

“The U.S. national debt now exceeds 100% of gross domestic product, crossing a once-unthinkable threshold, on the way toward breaking the record set in the wake of World War II.”

There’s some interesting choices of words in there, like “once-unthinkable” and “breaking the record”, phrases that seem to indicate danger.

If you feel nervous about the debt, here’s some things to consider. First, remember that personal debt and national debt don’t work the same way. Second, it’s important to take a longer term view when considering national debt.

A good example of a long term view is this chart of the U.K. government debt as a percentage of GDP. You can see the debt to GDP ration rose substantially above 100% at two major events in British history: the battles with France in the 19th century and the battles with Germany in the 20th century. Indeed, the ratio rose as high as 250% in the 20th century, only to subside to below 50% in the postwar era.

This is not to say that the ratio does not matter. It is to say that there is nothing magical about breaking the 100% ratio. The ratio could go higher. Chances are it will eventually go lower if for reasons of inflation if nothing else.

Writers — usually right wing writers — start bringing up the debt when left wing governments are in power. It’s a way for them to frighten people and pressure the government into spending less. (Such writers rarely bring up debt, especially for things like military expenditures, when the government is right wing.) Don’t be frightened. Governments have an obligation to tax and spend responsibility. They should not be basing their spending on this ratio.

(Image is from here.)

On running GPT-OSS-20B on my Macbook Pro

I’m been running gpt-oss-20b on my Macbook Pro for some time now. I typically run different LLMs on my Macbook using ollama as the basis for this. While I’ve tried various models, this model from OpenAI is the one I found that gave me the best answers to the questions I’ve been giving it. I even prefer the answers I get with gpt-oss-20b on my Mac over those I get from larger SaaS models provided by OpenAI, Anthropic and Google.

I’ve found it very good when it comes to asking it questions concerning lifestyle changes. For instance, I asked it for:

  • ideas on how to celebrate a big birthday
  • a 12 month plan to do Swedish death cleaning on my house
  • a 12 month plan to get my estate in order
  • a 12 month plan to get back in shape after a long layoff

I’ve found it not very good at all with questions around food and music:

  • coming up with music playlists based on certain criteria (over a third of the songs were hallucinations!)
  • coming up with spice mixes for cooking meats (the proportions were way off compared to tested recipes)

Like all LLMs, I found it works well as a means to generate ideas. It works ok or poorly when you start expecting exactitude in the answers. YMMV.

Other things I like about gpt-oss-20b running on ollama on my Macbook:

  • I can see it go through it’s reasoning when it comes up with the answer, which is a nice feature.
  • For some reason the default output format is one to my liking, and I can easily move it into Microsoft Word or Apple Pages for further formatting
  • I especially like it because I can ask personal questions and not have those being captured by a service in the cloud. It provides an extra layer of privacy.

I am running gpt-oss-20b and not gpt-oss-120b due to the capacity of my Macbook Pro. Indeed, even gpt-oss-20b takes up a fair amount of resources on my Macbook, despite it having a GPU to do the heavy lifting. I am not in a rush for these answers though, so it is fine. I’d run 120b if I could, but I have been happy with the response with 20b.

If you have a machine with a GPU and enough RAM and disk space, consider downloading ollama and this model and give it a go. Let me know what you discover!