Month: December 2024

Categories
AI technology

AI race erupts, gambling on startups, Google quantum chip, tips and help Bali street dogs

This is probably the last newsletter until Sunday, January 19, unless there is as much news in early January as there was last week. It is therefore an extra-long edition, which you can read (if you visit some of the links) over several days. 

Besides the usual attempt at news analysis, this time some tips for greater efficiency in the new year, a summary of last week's survey on this newsletter and a few reading and watching tips during Christmas. Finally, I draw your attention to a charity close to my heart: a contribution to the medical treatment of street dogs in Bali.

Google must hate OpenAI

It is no longer a coincidence: every time Google comes out with a new product, OpenAI launches something spectacular with which it attracts all the attention. Last week, Google launched Gemini 2, a so-called "Multimodal AI," or form of artificial intelligence that can process and integrate information from, for example, text, image, speech and video. 

The Verge highlights the interview with Google's AI front man Demis Hassabis, while VentureBeat published a detailed analysis. Gemini 2 is downright extraordinary, if not spectacular. Try it yourself by uploading a video, or sharing your screen live (for the brave ones among us), or your webcam. 

I used the webcam, asked and got tips on how to improve my workspace (rotate desk slightly and place a lamp from the front) and confirmation that the tree behind me in the garden is a banana tree. Generating time codes and captions to existing videos also worked flawlessly.

Marques Brownlee tested OpenAI' s Sora extensively. On X, he shared a number of videos he generated with Sora, with the news coverage in particular looking frighteningly good. His full review can be seen here. As usual with Brownlee, it is highly recommended viewing. Sora not only raises all sorts of ethical questions but appears to be possibly trained with game content, which could have legal implications.

Incidentally, sad news broke yesterday that 26-year-old Suchir Balaji, a whistleblower in the swelling wave of lawsuits against OpenAI over copyright violations, has taken his own life. 

A donkey with five legs

The launch of Sora took place during a 12-day period in which OpenAI presents something new every day, under the sad slogan "12 Days of Shipmas. Hopefully a name generated with AI, otherwise we should burn a candle this Christmas for the marketer who came up with it. 

Since OpenAI and Google speak so highly of their multimodal qualities and the initial tests with video are very encouraging, I expected them to have no problem with a test of a photo. Right? 

Image created with Midjourney.

I asked Google Gemini Advanced 2.0 Flash Experimental, another delightful name, if there was anything wrong with the above image I generated with Midjourney with the prompt "a man, a woman, a newborn baby and a donkey, two thousand years ago in the Middle East. 

Google Gemini 2.0 immediately saw that all sorts of things were wrong with the image:

"The image you have provided shows the traditional nativity scene, a religious representation of the birth of Jesus Christ. However, there are some historical inaccuracies in the image:

  • Palm tree: Palm trees are not native to Bethlehem, which is located in the Middle East. The climate in Bethlehem is more temperate and would not support the growth of palm trees.
  • Donkey: Although donkeys were common in ancient Judea, they were mainly used for work and transportation. It is less likely that a donkey would have been present in the stable during the birth of Jesus.
  • Architecture: The architecture of the buildings in the background may not be completely accurate for the period in which Jesus was born.
  • It is important to remember that the nativity scene is a symbolic representation and is not meant to be a historically accurate representation of the event."

Google and OpenAI both blind in plain sight

In itself it's clever that Google immediately recognized that it was a famous Biblical scene, though I had avoided all religious connotation in the prompt at Midjourney. Also commendable that it points out that there are few palm trees in Bethlehem and that the architecture may not be entirely accurate. I was waiting for Google to express doubts about the quality of the masonry and the shoddy paint job on the wall of the stable. 

What was especially striking: even after specifically asking several times if there was anything strange to see about the donkey, Google did not see that the poor animal has five legs. Likewise, ChatGPT 4o failed to see that this was a very special donkey; it noticed that only after specifically asking how many legs the donkey has. Midjourney generated more pathetic donkeys. It is typical of the state of AI: it is hit or miss.

Apple misfires 

Meanwhile, Apple released its flagship Apple Intelligence with very mediocre results for such a quality company. For example, one reporter's iPhone showed the Apple Intelligence-generated alert up: a
 that the New York Times reported that Israeli Prime Minister Netanyahu had been arrested. The BBC, according to Apple Intelligence, reported that the suspect in the murder of the United HealthCare CEO had committed suicide. 

Hallucination, or wishful thinking by Apple Intelligence?

Except: the New York Times and the BBC hadn't reported that at all. Oops. It would be good if the leading tech companies took a little more time to test their products. Now it seems like the first year of each product is basically the global beta test.

Good use of AI

Useful AI applications, despite the mistakes of the big players, are increasingly appearing on the market for specific purposes. For example, a traffic camera described as a "state-of-the-art Heads-Up machine", a name probably coined by the same genius behind the word Shipmas. Based on an analysis of the driving behavior of road users, I suspect at least swerving, or playing Diddy songs, the camera advises police to pull over certain drivers. 

Another application of AI that directly benefits the world is a program in the U.S. that results in a 21% higher chance of detecting breast cancer. AI is also being used successfully in England to detect forms of breast cancer that elude the human eye.

Throwing pudding against the wall

The challenge of tech startups is getting the technology right, which is hard enough, and then turning that technology into products for the right market. How difficult that is is confirmed by the disastrous results of the Tiger investment fund that have been leaked:

  • Invested in 315 startups in one year
  • Downgraded from $93 million to $65 million
  • Negative 15% IRR (performance in the bottom ten percent)

Those who invest in hundreds of startups in a year are basically just gambling, although spray and pray sounds nicer; it is the financial version of throwing pudding at the wall hoping something will stick. 

Cruise stops driving

Not coincidentally, there is a startup that has made betting, excuse me, prediction markets a business model: Polymarket. After the prediction market on Polymarket proved to be better at predicting the US presidential elections than sophisticated polling, CEO Shayne Coplan's star rose quickly. Coindex published this portrait of Coplan. In cryptocurrency circles, Polymarket is being hailed as an example of an application of crypto that for once is not aimed at replacing the current banking system.

Multinationals, of course, also gamble on startups. In 2016, General Motors bought three-year-old startup Cruise, which developed technology for self-driving cars, for a billion dollars. Then GM threw another $3.5 billion at it, only to pull the plug on it this week. Cruise founder Kyle Vogt was clear on X: "GM are a bunch of dummies." 

I have always followed Cruise closely, having turned down the opportunity to invest in Cruise's seed round through a syndicate nearly a decade ago. "Too early," was my infinitely dumb judgment. So, of course, within two years Cruise was bought for a billion by General Motors. Cruise illustrates how difficult it is to properly assess the value of startups: from a technology and product perspective, my judgment was correct that it could be another decade before self-driving cars are widely used. 

I did still get the congratulatory email, very nice.

Except that market maturity does not always determine the value of an innovation. GM cannot innovate and so Cruise's purchase was predictable. As an investor, I should have thought that way. As a big believer in self-driving cars and someone who does not own a car himself, I keep waiting for the joyous moment when cars become available on demand quicker than pizza's. For Cruise's team, it is unimaginably sad that a decade of work is effectively being thrown away.

Google announces revolutionary quantum chip

Researchers at Google have developed a new quantum chip named Willow, an important milestone in the quest for quantum computers. The experiment, described last week in Nature, shows that quantum computers, with the right error-correction techniques, can perform computations with increasing accuracy as they scale up, with the speed of this improvement crossing a crucial threshold. Current quantum computers are still too small and error-prone for most commercial or scientific applications.

The importance of this technology is phenomenally explained by Ray Harishankar: watch his accessible explanation in this panel at last year's ATX Singapore starting at minute 35:05. There is much discussion about how acute the danger is that quantum computers can decipher all passwords and crack Bitcoin. That it can be done in theory is clear; the question is when and how much computing power (and cost) it will take. Forbes came up with a solid analysis.

New app from Blogger, Twitter and Medium founder Ev Williams: Mozi

Each time it seems like there are too many social media outlets, a new one comes along. Most of them flop, of course, but Ev Williams previously founded Blogger, Twitter and Medium so his new app Mozi deserves attention. The idea of Mozi is that you can meet up with acquaintances in places where you might not expect each other. You actually share your future location.

In terms of marketing-speak, Mozi is just fine: it's described as "A Place For Your People" and its slogan reads "See You Sooner. Investor (at Google Ventures) M.G. Siegler, better known as Paris Lemon at Techcrunch with a pen like a razor, judges harshly: he doesn't like it. I like the idea, so if you want to find me and other acquaintances and friends on Mozi, this is the link. 

Freedom of speech on the Internet in 2024.

I had a problem with that other Ev Williams company, Medium. The company, which prides itself on being a medium for free expression, turned out to censor my newsletters because they contained links to Coinmarketcap, the cryptocurrency prices website. Medium simply refused to send out my newsletters because Coinmarketcap was allegedly used for spam. That's as absurd as refusing newsletters that link to the Weather Channel because you hate the weather forecast. The situation has since been remedied, but I started publishing on Substack.

Note the total lack of apology to a paying customer.

Survey results

There were far more responses to last week's survey than I expected, for which I thank you! The conclusions and recommendations are clear, but to make some sense of them it is important to remember that many readers of this newsletter have a specific profile.  

I still reach few students.

The percentage of startup founders, owners, CEOs and board members is disproportionately high (almost 25%) and the responses I receive via email also show a lot of interest in funding and investing. Only, among many other readers of this newsletter, there is much less interest in these topics. Although, of course, I still write based on what I noticed during the week, I do like to choose from these the topics that readers find interesting. That remains difficult, which is why the survey responses were very helpful:

  • over half of the readers who responded appreciate work tips that are useful to them 
  • for coverage of startups, there is much less interest (less than 50%)
  • there is a lot of interest in AI though, may even be highlighted more
  • there is enough coverage of crypto/Web3, no need for more
  • there is also sufficient focus on investing, no need for more 
  • only 12% would like to receive a newsletter more often during the week
  • more than half are interested in a podcast

Interestingly, from the "open suggestions" very different sounds were heard again: over a third of them asked for more attention to blockchain and tokenization. Of the readers who did not object to being mentioned, Stephan Fellinger asked for attention to passive income and Jeroen Verkroost suggested paying more attention to startups, especially in the field of AI, that could be helped by the readers: "That would be my definition of 'More information that is useful in your work'. And we all support entrepreneurs as possible. How nice is that."

Another suggestion that I'm definitely going to do something with is to ask a few questions weekly to specialists in my network. I want to make that interactive: what questions would you like to see answered? Mail me those questions and I'll put them to an expert in my network. 

Tip for higher efficiency in 2025

Let me take a recommendation from the survey to heart and share my most important work tip. This is not a smart app, difficult AI tool or savvy life hack. No; I haven't had a single notification turned on on my phone for over a decade. After going crazy from the red light indicating a new e-mail in the BlackBerry era, once I switched to an iPhone I decided to turn off as many notifications as possible.  

There's a nice tool online, the Trail Making Test, which shows right away how much harder it is to do tasks interchangeably. But you can also do this with a colleague or partner: 

  • clock how long it takes to write down numbers 1 through 26
  • Then measure how long it takes to write out the alphabet
  • add up the two times
  • Finally, measure how long it takes to write out 1a, 2b, 3c, 4d, 5b etc until you are at 26Z

I found the results baffling and think you will also be shocked.

The biggest misconception these days is that many people think you have to reply to messages immediately, or at least very quickly. That's not true. In fact, the unannounced phone call is almost extinct because it is not convenient when people who are communicating have to do so at the same time. Use the advantages of asynchronous communication and check your phone once every half hour. Your colleagues, your partner, friends, children, parents; no one cares if you answer after half an hour or an hour. We are all much less important than we think.

Reading and viewing tips for the holiday season

Here are some viewing and listening tips to go with the oliebollen:

  • A comprehensive look behind the scenes at OpenAI around the launch of ChatGPT two years ago; and a look ahead. Every company is moderately organized chaos, and a startup is by definition a gang trying to find its way. OpenAI is not any different.
  • Buy and read the debut novel by pleasant human and loyal reader of this newsletter, Nart Wielaard, about AI: Painless Barbed Wire. In this novel, Wielaard asks necessary questions about how artificial intelligence is getting a grip on the world and how humans are in danger of developing "machine-like" traits as a result.
  • Why AI is going to eat the browser. "For most of us, it’s hard to imagine life without an internet browser. But as AI disaggregates information from text, video, and music into unique remixable AI chatbot answer streams, it’s clear to me that over the next decade the browser will need to adapt or die." 
    A characteristically strong essay by former journalist, now investor, Om Malik.
  • Another great behind-the-scenes look at an extraordinary startup: OpenSea. Interesting read about a chaotic work environment, constantly changing priorities and problems with the SEC.  
  • I'm repeating myself, but want to emphasize that quantum computing is becoming very important, even if you think you won't have to deal with it: watch Ray Harishankar's explanation of the importance of quantum computing (starting at minute 35:05). Takes a few minutes and like me you may still not understand how it works, but you will understand why it is so important.

These are just two of the thousands of dogs Rhonda Lepsch has saved from a cruel death

Finally, my Christmas wish

This newsletter is and will remain free, because I write it to keep myself informed; and admittedly also from a lack of modesty, under the delusion that people will find it fun or interesting to read my opinion. 

I will never ask for tips, subscriptions or a donation for myself; but I do kindly ask you to support this special woman in paying her bill at the vet: Rhonda Lepsch, of JSP. You can follow Rhonda on Instagram here.

Rhonda is American, married to a Balinese and has lived in Bali for many years. Many people dump (literally) sick dogs with her knowing that Rhonda just cannot send dogs away. She gets them vaccinated, sterilized at her own expense, and takes in the sickest dogs herself that no one wants. No matter how crippled, blind, deaf or ugly they are: to Rhonda, every dog is equally special.

Help hundreds if not thousands of dogs

Not surprisingly, Rhonda is fantastic with dogs, I learned from her how to handle our own street dog Nala, whom we found at our doorstep last year more dead than alive.

At this point in time, Rhonda herself has 87 dogs, 25 of which live at her husband's dive center, which unfortunately went bankrupt during the Covid pandemic. She pays everything out of her own pocket.  

But in addition to taking care of those 87 dogs, Rhonda also pays for vaccinations and sterilizations of street dogs, and the costs of these are skyrocketing. Precisely this work is enormously important: of course it is good to adopt a street dog, as we did, but by supporting Rhonda you actually take care of many more dogs. Because vaccinated, healthy street dogs can have a good life in Bali. Unfortunately, unvaccinated, sick dogs have little chance and stand a chance of ending up as sate.

Right now, unfortunately, Rhonda and her street dogs are in trouble. She has helped so many street dogs this year that the outstanding bill at the excellent veterinarian Sunset Vet is already five thousand dollars. Therefore, Rhonda cannot have any more animals helped there. It means many sick street dogs on Bali will remain untreated in the middle of rainy season, with their lives at risk. 

You can support Rhonda and the dogs in Bali by helping pay the bill at Sunset Vet. Once she gets above zero, the balance is used by the clinic exclusively for medical assistance to new dogs brought by Rhonda. By supporting Rhonda, you are helping hundreds, if not thousands, of dogs.

This is how you can help

  • Click here for veterinarian Sunset Vet's donation form
  • very important: please enter "for JSP" in the description. Only then will your donation be used for Rhonda's account
  • choose the amount you want to donate
  • choose PayPal or a debit or credit card

My thanks are already great, regardless of the amount you donate. And if you don't donate anything, no problem, only I hope your gourmet set gives out on Christmas Eve.

I wish you and all your loved ones all the happiness, love and health in 2025! Thank you for your interest and see you next year.

Categories
AI investing crypto NFTs technology

Bitcoin over $100,000, but still beaten by Ethereum and Nvidia

Please participate as happily as this reader and complete the short survey about this newsletter. Who knows, maybe it will help! Image: generated with Midjourney.

This week I'm asking for your help via a survey about this newsletter. Also: a look at Bitcoin's jump above the $100,000 mark, Ethereum and the start of altcoin season, Hawk Tuah girl is in trouble, AI applications to watch out for and crypto gifts for the holidays.

You ask, we deliver 

Well, maybe 😉 After almost two years and eighty newsletters with what I used to send to friends and colleagues completely subjectively, it's time to gauge which topics are found most and least interesting. Therefore, I invite you to fill out this short survey, which takes less than two minutes. The main results will be shared next week. The last question allows for suggestions, comments and statements. The survey is anonymous, but if you appreciate it, I am happy to include your name and LinkedIn profile in valuable feedback.

Bitcoin over the $100,000 mark and back again

It is always enjoyable when something huge happens in an area that the BBC heartily dislikes but has to report on. Think of Max Verstappen winning the world title and thwarting Briton Lewis Hamilton's record.

With the face of a spoiled child being served Brussels sprouts and chicory, the reporter filed this report, about the moment Bitcoin became worth more than $100,000 this week. Most media only report the investment risks, but have been completely overlooking some crucial elements of Bitcoin for years.

What is unique about Bitcoin is that it is a completely decentralized network with no central party controlling the number of coins in circulation, so unlike a system with a central bank. As a result, there is also no "crypto-bank account", but you have complete control over your own assets.

How important that is, Netscape founder and investor Marc Andreessen explained in Joe Rogan's podcast through a detailed explanation of the concept of being "debanked;" what happens when you lose your bank accounts and credit cards as a person or company without any explanation. This has happened to many entrepreneurs in the fintech and crypto world in recent years. More and more people worldwide, and not just in undemocratic countries, therefore value self custody of their assets.

Big difference between 2018 and 2024

The often-quoted brilliant Jennifer Zhu Scott recalled last week a panel she participated in at the 2018 World Economic Forum in Davos, when Bitcoin first took its place on the main stage. Zhu Scott debated with Nobel prize winner Robert Shiller and a top executive from Sweden's central bank:

"At the time, coming out openly for Bitcoin was a career risk. But I believed in the ideal. I understood its powerful implications for the world and chose to champion it. Many of the ideas I shared six years ago have become reality.

  • Bitcoin has disrupted gold more than the dollar.
  • Smaller countries have begun to include Bitcoin in their national reserves.
  • Bitcoin wasn't going anywhere-it embodied the ideal of decentralization, and that ideal remains incredibly powerful.

While the rise of Bitcoin has been extraordinary, the ecosystem has evolved in ways I had not fully anticipated:

  • Decentralization vs. Centralization: Bitcoin was born from the ideal of decentralization. Today, however, the ecosystem is becoming increasingly centralized, with figures like Michael Saylor exerting excessive influence.
  • The Obsession with Price: Six years ago, I argued that price was the least important aspect of Bitcoin; its real strength lay in decentralization. Yet Bitcoin is now seen as an asset and the conversation is almost entirely about price movements.

I am proud to have been one of the early pioneers in this field. Thanks to Satoshi Nakamoto's revolutionary vision, we have witnessed the birth of countless breathtaking projects aimed at enabling scalable decentralization. In an era where AI is consuming our data, creating deepfakes and undermining trust, I am grateful for the emergence of Web3 technologies-those offering solutions such as data sovereignty, immutability and authentication. These tools provide a counterbalance in a world increasingly defined by unchecked digital power.

To those who rejected the picture I painted eight years ago: I smile today. Bitcoin's journey is far from over, and this milestone is just the beginning of what is possible."

From Davos to Hong Kong

The debate, in which things got especially feisty between Zhu Scott and Schiller, can be seen here. In retrospect, it is downright scandalous to see how crypto was portrayed in a certain light at this conference, where supposedly free exchange of ideas takes place.

First of all, the panel was called "The Crypto Asset Bubble." That is especially hilarious when you consider that the price of Bitcoin at that time was $11,000; so anyone who had bought Bitcoin at that time would have seen that investment increase in value nearly tenfold within seven years. Do me a bubble like that more often!

But the text with which the moderator introduces Bitcoin really defies all standards of decency: "Bitcoin emerges from the world of nerds and criminals." As a half nerd, I then always think, "what if you replace the word 'nerd' here with Jews or Asians?" Suddenly the same sentence is a lot less acceptable.

By the way, 2018 was the last time I attended the World Economic Forum. Next month, the media will again report in full on the party in the mountains, which I wrote about earlier:

"Participants reported that the number of women they encountered in Davos was as high as the number of MMA fighters at the annual Women's Hairdressing Day. Women are almost as rare during WEF as dark-skinned people. Like motorcyclists on a Sunday ride or penguins in a zoo, I caught myself in Davos politely waving back or nodding to other fellow pigmented people.

A week later, no WEF participant can remember what else was discussed or agreed upon, because unlike the COP climate conferences, for example, Davos is not about jointly formulating measurable goals. There is old-fashioned networking and job hunting.

Because I am sorry to disappoint the conspiracy thinkers, but there is no talk at WEF of world domination by a small, ruling elite at the expense of the common people; there is not much thought about the future at all. WEF excels mainly in zizagging into the future, looking in the rearview mirror - with glasses dipped in the cheese fondue."

In fairness, the "Swiss network effect" was very useful to me personally. I got to know Zhu Scott just before WEF at a crypto conference in St. Moritz, after which I concluded through her appearance on the panel with Schiller that she has a special gift of being able to position her immense knowledge of even the most intricate technical details, within global trends and developments.

Zhu Scott was kind enough to spend a few hours in Hong Kong a few months later to tutor this rather useless Dutchman on decentralization and blockchain. In my experience, this accessibility and willingness to share knowledge is more often seen in the crypto world than in the traditional IT sector, not to mention the financial world.

No one expects it to happen again, but over the last five years Ethereum has been a much better investment than Bitcoin and even better than Nvidia.

Ethereum did much better than Bitcoin

By investors worldwide, including those from the traditional financial industry, Bitcoin has been accepted this year as an investment product that is best compared to investing in gold: it is not used as a means of payment, but as a long-term investment. That is why it remains striking that if you look at the market with a view over five years instead of five minutes, Ethereum has appreciated more than twice as much as Bitcoin.

Indeed, if we compare the performance of the two largest crypto currencies with that of the three most valuable companies in the world, it appears that Ethereum has even outperformed stock market darling Nvidia. That's not saying it will be the case again in the next five years, but it's still interesting to keep in mind.

Christmas season? It's altcoin season

When Bitcoin's price rise has been very rapid, as it has been in recent months, it is usually a matter of time until some profit taking happens and the proceeds from that sale of Bitcoin are invested in other cryptocurrencies: altcoins.

That happened last week, when many altcoins peaked, including BNB, Dogecoin (DOGE), XRP and Chainlink (LINK). In the previous cycle, in 2021, altcoins outperformed Bitcoin for nearly five consecutive months.

In the crypto world, it's not Christmas, but altcoin season. Source: Coinmarketcap.

The question, of course, is which sectors within cryptocurrencies will benefit the most from the upcoming altcoin season. To the frustration of all sincere blockchain developers, at the moment it seems especially all meme coins, all crazy coins without any underlying value, will benefit the most.

Hawk Tuah Girl goes crypto

She was world famous for a few days this summer: Haliey Welch, better known as hawk tuah girl. Last week she was accused of involvement in large-scale scams by releasing a coin called HAWK that became worth nearly half a billion dollars in a short period of time, after which it collapsed completely(91% drop).

Now, few will have expected the same stringency of monetary policy from Ms. Welch as from Alan Greenspan, but this was pretty ugly. Stephen Findeisen, better known as YouTuber Coffeezilla, was quick toexpose the so-called  rug pull. But also completely unknown or random people can issue tokens, even kids.

Teenager makes meme coins

"On the evening of November 19, art adviser Adam Biesk was finishing work at his California home when he overheard a conversation between his wife and son, who had just come downstairs. The son, a kid in his early teens, was saying he had made a ton of money on a cryptocurrency that he himself had created.

At first, Biesk paid little attention to it. He knew his son was experimenting with crypto, but the thought that he had made a small fortune before bedtime seemed far too unlikely. "We didn't actually believe it," Biesk says. But when the phone started ringing incessantly and his wife was inundated with angry messages on Instagram, Biesk realized that his son was telling the truth-albeit not quite the full story."

In an excellent article, Ars Technica explains how and why Pump.Fun appears to be a new home for anyone looking to quickly release their own meme coin. The method is simple: the potential profits with meme coins are always coupled with immeasurable risks. Only suitable for the gambler with a strong stomach.

Interesting crypto categories

As every week, I want to emphasize that I am not providing investment advice, but these appear to be interesting categories of cryptocurrencies:

  • Real world assets (RWA): These are cryptoprojects that represent tangible, physical objects, such as gold, real estate, art or commodities that are "tokenized." By converting these into digital tokens, they can be more easily traded and managed, while ownership is transparent and verifiable in the blockchain.
  • AI tokens (which may or may not actually have anything to do with AI): This refers to crypto projects that capitalize on the theme of artificial intelligence, but do not always actually incorporate substantial AI technology. Some projects attempt to actually integrate AI applications, such as "smart contracts" with machine learning functionality, while others simply use the term "AI" from a marketing standpoint.
  • DePin or DePIN (Decentralized Physical Infrastructure Network): This is a blockchain-based model for decentralized management of physical infrastructure. This could include networks for Internet access, energy or mobility, in which the infrastructure is not controlled by one central body. Imagine: Uber with only drivers and customers, with no central organization taking 20-30% of revenue.

Venture capital firm Andreessen Horowitz published a nice overview yesterday of what it believes to be promising sectors in 2025.

Short news

Tesla is buoyed by Musk's bromance with President Trump, but Ethereum also rose strongly this week.
  • Tesla, Ethereum, Amazon and Meta are the winners of the stock market week, which actually had no losers. Yes, the S&P500 lagged behind tech stocks and crypto stocks, but the risk profile is also lower. At 28% up, the S&P500 is having a phenomenal year. Especially compared to savings accounts...

It's a funny contest run by AI experts and scientists: get an AI bot to declare its love to you and win thousands of dollars.

"It can be helpful to agree on a 'proof of humanity' word that your trusted contacts can ask you," Near wrote. "Should a strange and urgent voice or video call from you come in, this can help them confirm that they are really talking to you and not a deepfake/deepcloned version of you."

When someone suggested this on X last year it may have seemed far-fetched, but now the FBI advises  families to put together their own secret password, or security phrase. When scammers and criminals then try to impersonate a member of the family over the phone, for example a supposedly lost child asking her parents to send her money quickly, the family password or security phrase can be used to verify who is on the line. My guess is that many families will choose to use phrases from popular movies as security phrases, such as from Harry Potter or Home Alone.

As many as 87% of all startups that participated in Y Combinator's program, which ends with a pitch to investors, were AI-related in some way. But while the focus is mostly on consumer applications, the greatest use of AI is actually taking place within enterprise environments. Techcrunch advises companies to look closely at CTGT, Galini, Raycaster and HumanLayer.

Most analysts overestimate energy demand and underestimate technological advances. That is the gist of an argument by The Economist, which has looked at estimates of the global cost of an energy transition to a carbon-free world made by various economists, consultants and other researchers-the kind of estimates that are routinely used as the basis for policy.

These estimates range from about $3 trillion a year to nearly $12 trillion a year, huge sums. But these figures are grossly exaggerated. The good news, according to The Economist, is that the energy transition is going to be many times cheaper.

I don't claim to put all of the suggested products on my wish list myself, but you'll agree that this description of a Christmas gift does grab attention: "the Chipped Social Nail Set includes NFC-equipped nails that can connect to blockchain experiences and crypto wallets, creating an interactive way to show off your crypto style. It's the perfect gift for the bold, stylish crypto fanatic in your life." How did we ever live without NFC nails connecting to your crypto wallet?

I close with once again the kind request to complete the short questionnaire about this newsletter. Thank you very much in advance!

Categories
AI invest crypto technology

The big trends of 2024: AI, crypto and carbon removal

There are currently three major trends in technology driven by technological, as well as sociological and political currents: AI, crypto and carbon removal. These groundbreaking developments, like any major innovation, are received with skepticism, a pattern that has been evident for decades.

PC: "too expensive and useless"

In the 1980s, when the personal computer emerged, personal computers were mostly seen as too expensive for a device without many relevant applications. That quickly changed thanks to price reductions and standardization of software, after MS-DOS became the world standard thanks to a sophisticated licensing model by Microsoft. The word processor and spreadsheet quickly made the PC indispensable in the office.

Internet: "too difficult and dangerous"

In the 1990s, this pattern repeated itself with the Internet. The personal computer was seen as a work tool, not a potential mass medium. Bill Gates even declared that the Internet suffered from lack of standards, it was insecure and far too complicated, which is why he did not use the word Internet even ten times in his book The Road Ahead.

Bill preferred to talk about the information super highway, which he was going to build himself with the closed MSN, which we never heard anything more about. Yet within a few years, email, the Web browser and applications such as eBay, Amazon and Google made the Internet accessible to consumers.

In the Netherlands, it took until late 1996 for the NOS Journaal to understand that the Internet was about to become a serious mass medium, although Joop van Zijl still compared computer penetration to that of the microwave oven.

Smartphones: "only for representatives"

When the iPhone hit the market in 2007, the Blackberry reigned supreme in the business market. Although most of the population in developed countries already had a cell phone, often a Nokia, criticism of the iPhone was not muted. "Too expensive, only useful for sales representatives," was the verdict of a friend from the world of IT. Incidentally, the same chap who ten years earlier judged the cell phone as "only useful for drug dealers," a common sentiment.

Microsoft CEO Steve Ballmer laughed off the iPhone in a video in which, as he was taught by PR people, he quickly switched to promoting the company's own Windows Mobile which we also never heard anything more about. It makes CEO Satya Nadella's feat of completely revitalizing Microsoft after Ballmer all the more galling, but about that another time.

AI, crypto and carbon removal on the turn

Right now we are seeing the exact same patterns as before, but now about AI, crypto and carbon removal:

  • AI is often dismissed as useful for work, but without useful applications for consumers.
  • Crypto is criticized with comments like, "Name an application." Meanwhile, the first application lies in something as basic as redesigning the banking system, with each user managing their own account and making banks obsolete. Apparently, the significance of this is missed by many. Tip: Never get into an argument with people who were too lazy to read the Bitcoin white paper but have an opinion.
  • Carbon removal is often characterized as a fraud, referring to familiar examples such as inefficient cooking ovens, without knowing or understanding the complexity and potential of projects that do actually remove carbon from the atmosphere, such as ocean fertilization. This kind of removal of carbon from the atmosphere is the biggest task facing the world in the coming decades. Tip: Never engage in climate change discussions with people who were too lazy to read the summary of recent IPCC reports.

Admittedly, I have a personal fascination with how innovations break through or fail. That's why both my 1993 graduate thesis and my 2001 book were both called "In Search of the Holy Grail," although some weirdo photoshopped the cover of my book which, by the way, is still on sale in large numbers. And not because of its great success.

I learned more from Megamistakes than Megatrends. Everyone knows Rodgers' adoption curve, but it remains mysterious why one innovation catches on and another flops mercilessly. For carbon removal, crypto and AI, there are several key success factors, some of which I want to highlight.

CO2 success was not during COP29

Breakthroughs in carbon removal require political will. All media were focused on the COP29 climate summit in Baku, but in the meantime, successes were being made in Brussels and Washington in the fight against climate change.

In Brussels, the European Council approved the creation of the first EU-wide certification framework for permanent carbon removal, carbon farming and carbon storage in products. This voluntary framework is intended to create a certification system that can quantify, monitor and verify carbon removals and counteract greenwashing; carbon farming. The EU's adoption of the new rules marks the last major legislative step to give the green light to the creation of the new certification framework for carbon removal.

Now in Dutch: standards are being introduced that will allow companies and citizens to actually offset their carbon emissions, and not by planting or preserving flimsy forests, but by measurably reducing CO2 emissions or even better, removing CO2 from the atmosphere.

Democrats and Republicans together for carbon removal

In the United States, a bill was introduced by Senators Lisa Murkowski (Republican, Alaska) and Michael Bennet (Democrat, Colorado) seeking to expand carbon removal subsidies for a wide range of technologies intended to permanently remove carbon dioxide from the air and seas.

The bill is unlikely to be passed by the current Congress yet due to time constraints, but its introduction indicates that subsidies for carbon removal will be expanded even under President Trump. The fact that the bill was introduced by senators from both parties, a rarity these days, is hopeful.

AMCs for CO2

In coming years, watch for the term Advanced Market Commitment (AMC), explained here by the Economist: no matter how the political winds blow, the pressure from society for decarbonization is so great that smarter companies are independently seeking to remove or minimally offset their own carbon footprint, by funding techniques that remove carbon for the long term; preferably forever. Salesforce, Google, Meta and Microsoft are just the first from a long list of companies that will fund AMCs.

As another example, it was announced last week that Planetary Technologies has removed 138 tons of CO2 through "Ocean Alkalanity Enhancement (OAE)," which, by adding minerals or substances, increases alkalinity, the ocean's capacity to absorb CO2e, with the goal of sequestering CO₂ and combating climate change. Buyers of the associated carbon removal credits were Shopify (96 tons) and Stripe (42 tons) under a "pre-purchase agreement. In Scrabble, you don't put it easily, but it really exists and will be used a lot.

Old school tech compared to AI and crypto

Stock market valuations are a reflection of market expectations, and the enthusiasm around AI and crypto shows that investors have confidence in their longer-term potential. I have created four virtual "baskets" that I have posted about before:

  • 'MANAAM': the old school tech companies
  • Spotlight 9: the nine I believe to be leading tech investments
  • AI Spotlight 9: nine companies benefiting from AI
  • Crypto Spotlight 9: the biggest nine cryptos measured by market value

Old school tech MANAAM: +36%

In the broader tech sector, established players continue to dominate. At one time investors were fans of the term FANG (for Facebook, Apple, Netflix and Google, as if Microsoft meant nothing), but let's take the "MANAAM" group consisting of Meta (formerly Facebook), Apple, Microsoft, Amazon, Alphabet (formerly Google) and Netflix. The average increase in shares of this now classic little club this year is a whopping 35.9%. That's phenomenal from an investment perspective, until you consider that the S&P 500 is also up 27.19% this year.

Spotlight 9: +63%

Microsoft(14%), Alphabet(22.28%) and Apple(27.84%) are not even outperforming the index. While investors buy tech stocks for the higher price appreciation, compensating for the higher risk.

Not a buy recommendation, but indicative: the Spotlight 9 is +63%

However, those who had bought the Spotlight 9, which consists of the major tech companies and the two largest crypto currencies Bitcoin (+119%) and Ethereum (+57%), would have already seen their investment portfolio rise 63.37% this year. Compared to the MANAAM, Netflix is missing from the Spotlight 9, while Nvidia (+187%) has obviously been added as the world's most valuable technology company.

AI Spotlight 9: +76%

The valuation of AI-driven companies such as Nvidia, which play a key role in the development of AI infrastructure, has reached record highs. This shows that the market recognizes the speed at which these AI-powered companies are seeing their results soar.

Despite AMD, Gigabyte and Super Micro, the AI Spotlight 9 does as much as + 76%

Since Nvidia is already included in the Spotlight 9, I left out the market leader in my also completely arbitrary "AI Spotlight 9," consisting of nine companies that I suspect AI will allow them to grow faster than the leading large tech companies (the MANAAM group) and perhaps even faster than the Spotlight 9.

With 76.11% growth, that is certainly the case this year, with it being entirely remarkable that this increase came about despite Super Micro (which saw the auditor go the distance), AMD (-1%) and Gigabyte, hardware parties that did not keep up with the growth of the rest. Software company Palantir (+305%), which I wrote about in early November, more than makes up the difference.

Crypto Spotlight 9: +191%

Since the approval earlier this year of Bitcoin ETFs, tens of billions have already flowed from the traditional investment world toward crypto. The wait was for the moment when the "alt rotation" would begin, the moment when more money flows into other cryptocurrencies than Bitcoin, which counts as the unofficial kickoff of "altcoin season. That moment occurred yesterday, when the Ethereum Spot ETF net inflows, outpaced those to Bitcoin.

Crypto Spotlight 9: +191% and this does not include memecoin.

So the real daredevil is now stepping big into the craziest coins that often have no underlying value at all, but that is as risky as putting everything on red or black in a casino. A less risky strategy, insofar as that is possible in crypto, is to spread out in the biggest cryptocurrencies and take advantage of overall sentiment.

The "Crypto Spotlight 9" consists of the largest crypto currencies measured by market value, excluding stable coins, memecoins (crypto giblets) and tokens linked to crypto exchanges such as BNB.

That group, listed alphabetically as Avalanche, Bitcoin, Cardano, Ethereum, Solana, Stellar, Toncoin, TRON and XRP, achieved a 191% increase so far this year. So is this a buy recommendation? Absolutely not.

What I do recommend to anyone active in technology and innovation is to look into AI, carbon removal technology, blockchain and crypto-currencies. Just like in the 1980s with the personal computer, the Internet in the 1990s and the smartphone 15 years ago, these are developments that are unstoppable worldwide.

A practical way to stay informed is to then invest a bit in those sectors, with my advice being to do so only with money you don't need for rent, mortgage or other daily concerns. Even within technology and crypto, it certainly pays to look closely at what the intended investments actually involve; what does Palantir actually do, is Ethereum threatened by Solana and SUI; and isn't it funny to take a small gamble on memecoins after all?

Anyone who puts in some money will start to inform themselves. The alternative is to write a weekly newsletter about tech and innovations, but that also requires a huge ego.

Warm regards, thanks for your interest and see you next week!