Technology

EconTalk Extra on Erik Hoel

Joy BuchananLeave a comment

Sometimes a Russ Roberts podcast gets an “Extra” post following up on the topic. I wrote an Extra for the Erik Hoel on the Threat to Humanity from AI episode:

Who’s Afraid of Artificial Intelligence? is the title of my Extra

Hoel’s main argument is that if AI becomes more intelligent than humans, it could pose a serious threat. What if the AI agents start to treat humans the way we currently treat wild deer, not necessarily with malice but without much regard for the welfare of every human individual?

Things that are vastly more intelligent than you are really hard to understand and predict; and the wildlife next door, as much as we might like it, we will also build a parking lot over it at a heartbeat and they’ll never know why. 

County Demographic Data: A Clean Panel 1969-2023

James Bailey2 Comments

Whenever researchers are conducting studies using state- or county-level data, we usually want some standard demographic variables to serve as controls; things like the total population, average age, and gender and race breakdowns. If the dataset for our main variables of interest doesn’t already have this, we go looking for a new dataset of demographic controls to merge in; but it has always been surprisingly hard to find a clean, easy-to-use dataset for this. For states, I’ve found the University of Kentucky’s National Welfare Database to be the best bet. But what about counties?

I had no good answer, and the best suggestion I got from others was the CDC SEER data. As so often, the government collected this impressively comprehensive dataset, but only releases it in an unusable format- in this case only as txt files that look like this:

I cleaned and reformatted the CDC SEER data into a neat panel of county demographics that look like this:

I posted my code and data files (CSV, XLSX, and DTA) on OSF and my data page as usual. I also posted the data files on Kaggle, which seems to be more user-friendly and turns up better on searches; I welcome suggestions for any other data repositories or file formats you would like to see me post.

HT: Kabir Dasgupta

Triumph of the Data Hoarders

James Bailey3 Comments

Several major datasets produced by the federal government went offline this week. Some, like the Behavioral Risk Factor Surveillance Survey and the American Community Survey, are now back online; probably most others will soon join them. But some datasets that the current administration considers too DEI-inflected could stay down indefinitely.

This serves as a reminder of the value of redundancy- keeping datasets on multiple sites as well as in local storage. Because you never really know when one site will go down- whether due to ideological changes, mistakes, natural disasters, or key personnel moving on.

External hard drives are an affordable option for anyone who wants to build up their own local data hoard going forward. The Open Science Foundation site allows you to upload datasets up to 50 GB to share publicly; that’s how I’ve been sharing cleaned-up versions of the BRFSS, state-levle NSDUH, National Health Expenditure Accounts, Statistics of US Business, and more. If you have a dataset that isn’t online anywhere, or one that you’ve cleaned or improved to the point it is better than the versions currently online, I encourage you to post it on OSF.

If you are currently looking for a federal dataset that got taken down, some good places to check are IPUMS, NBER, Archive.org, or my data page. PolicyMap has posted some of the federal datasets that seem particularly likely to stay down; if you know of other pages hosting federal datasets that have been taken down, please share them in the comments.

After the Fall: What Next for Nvidia and AI, In the Light of DeepSeek

Scott BuchananLeave a comment

Anyone not living under a rock the last two weeks has heard of DeepSeek, the cheap Chinese knock-off of ChatGPT that was supposedly trained using much lower resources that most American Artificial Intelligence efforts have been using. The bearish narrative flowing from this is that AI users will be able to get along with far fewer of Nvidia’s expensive, powerful chips, and so Nvidia sales and profit margins will sag.

The stock market seems to be agreeing with this story. The Nvidia share price crashed with a mighty crash last Monday, and it has continued to trend downward since then, with plenty of zig-zags.

I am not an expert in this area, but have done a bit of reading. There seems to be an emerging consensus that DeepSeek got to where it got to largely by using what was already developed by ChatGPT and similar prior models. For this and other reasons, the claim for fantastic savings in model training has been largely discounted. DeepSeek did do a nice job making use of limited chip resources, but those advances will be incorporated into everyone else’s models now.

Concerns remain regarding built-in bias and censorship to support the Chinese communist government’s point of view, and regarding the safety of user data kept on servers in China. Even apart from nefarious purposes for collecting user data, ChatGPT has apparently been very sloppy in protecting user information:

Wiz Research has identified a publicly accessible ClickHouse database belonging to DeepSeek, which allows full control over database operations, including the ability to access internal data. The exposure includes over a million lines of log streams containing chat history, secret keys, backend details, and other highly sensitive information.

Shifting focus to Nvidia – – my take is that DeepSeek will have little impact on its sales. The bullish narrative is that the more efficient algos developed by DeepSeek will enable more players to enter the AI arena.

The big power users like Meta and Amazon and Google have moved beyond limited chatbots like ChatGPT or DeepSeek. They are aiming beyond “AI” to “AGI” (Artificial General Intelligence), that matches or surpasses human cognitive capabilities across a wide range of cognitive tasks. Zuck plans to replace mid-level software engineers at Meta with code-bots before the year is out.

For AGI they will still need gobs of high-end chips, and these companies show no signs of throttling back their efforts. Nvidia remains sold out through the end of 2025. I suspect that when the company reports earnings on Feb 26, it will continue to demonstrate high profits and project high earnings growth.

Its price to earnings is higher than its peers, but that appears to be justified by its earnings growth. For a growth stock, a key metric is price/earnings-growth (PEG), and by that standard, Nvidia looks downright cheap:

Source: Marc Gerstein on Seeking Alpha

How the fickle market will react to these realities, I have no idea.

The high volatility in the stock makes for high options premiums. I have been selling puts and covered calls to capture roughly 20% yields, at the expense of missing out on any rise in share price from here.

Disclaimer: Nothing here should be considered as advice to buy or sell any security.

DeepSeek vs. ChatGPT: Has China Suddenly Caught or Surpassed the U.S. in AI?

Scott Buchanan1 Comment

The biggest single-day decline in stock market history occurred yesterday, as Nvidia plunged 17% to shave $589 billion off the AI chipmaker’s market cap. The cause of the panic was the surprisingly good performance of DeepSeek, a new Chinese AI application similar to ChatGPT.

Those who have tested DeepSeek find it to perform about as well as the best American AI models, with lower consumption of computer resources. It is also available much cheaper. What really stunned the tech world is that the developers claimed to have trained the model for only about six million dollars, which is way, way less than the billions that a large U.S. firm like OpenAI, Google, or Meta would spend on a leading AI model. All this despite the attempts by the U.S. to deny China the most advanced Nvidia chips. The developers of DeepSeek claim they worked with a modest number of chips, models with deliberately curtailed capacities which met U.S. export allowances.

One conclusion, drawn by the Nvidia bears, is that this shows you *don’t* need ever more of the most powerful and expensive chips to get good development done. The U.S. AI development model has been to build more, huge, power-hungry data centers and fill them up with the latest Nvidia chips. That has allowed Nvidia to charge huge profit premiums, as Google and other big tech companies slurp up all the chips that Nvidia can produce. If that supply/demand paradigm breaks, Nvidia’s profits could easily drop in half, e.g., from 60+% gross margins to a more normal (but still great) 30% margin.

The Nvidia bulls, on the other hand, claim that more efficient models will lead to even more usage of AI, and thus increase the demand for computing hardware – – a cyber instance of Jevons’ Paradox (where the increase in the efficiency of steam engines in burning coal led to more, not less, coal consumption, because it made steam engines more ubiquitous).

I read a bunch of articles to try to sort out hype from fact here. Folks who have tested DeepSeek find it to be as good as ChatGPT, and occasionally better. It can explain its reasoning explicitly, which can be helpful. It is open source, which I think means the code or at least the “weights” have been published. It does seem to be unusually efficient. Westerners have downloaded it onto (powerful) PCs and have run it there successfully, if a bit slowly. This means you can embed it in your own specialized code, or do your AI apart from the prying eyes of ChatGPT or other U.S. AI providers. In contrast, ChatGPT I think can only be run on a powerful remote server.

Unsurprisingly, in the past two weeks DeepSeek has been the most-uploaded free app, surpassing ChatGPT.

It turns out that being starved of computing power led the Chinese team to think their way to several important innovations that make much better use of computing. See here and here for gentle technical discussions of how they did that. Some of it involved hardware-ish things like improved memory management. Another key factor is they figured out a way to only do training on data which is relevant to the training query, instead of training each time on the entire universe of text.

A number of experts scoff at the claimed six million dollar figure for training, noting that if you include all the costs that were surely involved in the development cycle, it can’t be less than hundreds of millions of dollars. That said, it was still appreciably cheaper than the usual American way. Furthermore, it seems quite likely that making use of answers generated by ChatGPT helped DeepSeek to rapidly emulate ChatGPT’s performance. It is one thing to catch up to ChatGPT; it may be tougher to surpass it. Also, presumably the compute-efficient tricks devised by the DeepSeek team will now be applied in the West, as well. And there is speculation that DeepSeek actually has use of thousands of the advanced Nvidia chips, but they hide that fact since it involved end-running U.S. export restrictions. If so, then their accomplishment would be less amazing.

What happens now? I wish I knew. (I sold some Nvidia stock today, only to buy it back when it started to recover in after-hours trading). DeepSeek has Chinese censorship built into it. If you use DeepSeek, your information gets stored on servers in China, the better to serve the purposes of the government there.

Ironically, before this DeepSeek story broke, I was planning to write a post here this week pondering the business case for AI. For all the breathless hype about how AI will transform everything, it seems little money has been made except for Nvidia. Nvidia has been selling picks and shovels to the gold miners, but the gold miners themselves seem to have little to show for the billions and billions of dollars they are pouring into AI. A problem may be that there is not much of a moat here – – if lots of different tech groups can readily cobble together decent AI models, who will pay money to use them? Already, it is being given away for free in many cases. We shall see…

Free Webinar, Jan. 25: Practical and Ethical Aspects of Future Artificial Intelligence

Scott BuchananLeave a comment

As most of us know, artificial intelligence (AI) has taken big steps forward in the past few years, with the advent of Large Language Models (LLM) like ChatGPT. With these programs, you can enter a query in plain language, and get a lengthy response in human-like prose. You can have ChatGPT write a computer program or a whole essay for you (which of course makes it challenging for professors to evaluate essays handed in by their students).

However, the lords of Big Tech are not content. Their goal is to create AI with powers that far surpass human intelligence, and that even mimics human empathy. This raises a number of questions:

Is this technically possible? What will be the consequences if some corporations or nations succeed in owning such powerful systems? Will the computers push us bumbling humans out of the way? Will this be a tool for liberation or for oppression? This new technology coming at us may affect us all in unexpected ways. 

For those who are interested, there will be a 75-minute webinar on Saturday, January 25 which addresses these issues, and offers a perspective by two women who are leaders in the AI field (see bios below). They will explore the ethical and practical aspects of AI of the future, from within a Christian tradition. The webinar is free, but requires pre-registration:

https://network.asa3.org/events/EventDetails.aspx?id=1889760&group=

Here are bios of the two speakers:

Joanna Ng is a former IBM-er, pivoted to a start-up founder, focusing on Artificial Intelligence, specialized in Augmented Cognition, by integrating with IoT and Blockchain, in the context of web3, by applying design-thinking methodology. With forty-nine patents granted to her name, Joanna was accredited as an IBM Master Inventor. She held a seven-year tenure as the Head of Research, Director of the Center for Advanced Studies, IBM Canada. She has published over twenty peer-reviewed academic publications and co-authored two computer science books with Springer, The Smart Internet, and The Personal Web. She published a Christianity Today article called “How Artificial Intelligence Is Today’s Tower of Babel” and published her first book on faith and discipleship in October 2022, titled Being Christian 2.0.

Rosalind Picard is founder and director of the Affective Computing Research Group at the MIT Media Laboratory; co-founder of Affectiva, which provides Emotion AI; and co-founder and chief scientist of Empatica, which provides the first FDA-cleared smartwatch to detect seizures. Picard is author of over three hundred peer-reviewed articles spanning AI, affective computing, and medicine. She is known internationally for writing the book, Affective Computing, which helped launch the field by that name, and she is a popular speaker, with a TED talk receiving ~1.9 million views. Picard is a fellow of the IEEE and the AAAC, and a member of the National Academy of Engineering. She holds a Bachelors in Electrical Engineering from Georgia Tech and a Masters and Doctorate, each in Electrical Engineering and Computer Science, from MIT. Picard leads a team of researchers developing AI/machine learning and analytics to advance basic science as well as to improve human health and well-being, and has served as MIT’s faculty chair of their MindHandHeart well-being initiative.

Study Shows AI Can Enable Information-Stealing (Phishing) Campaigns

Scott BuchananLeave a comment

As a computer user, I make a modest effort to stay informed regarding the latest maneuvers by the bad guys to steal information and money. I am on a mailing list for the Malwarebytes blog, which publishes maybe three or four stories a week in this arena.

Here are three stories from the latest Malwarebytes email:

 ( 1 )   AI-supported spear phishing fools more than 50% of targets A controlled study reveals that 54% of users were tricked by AI-supported spear phishing emails, compared to just 12% who were targeted by traditional, human-crafted ones. ( 2 )  Dental group lied through teeth about data breach, fined $350,000 Westend Dental denied a 2020 ransomware attack and associated data breach, telling its customers that their data was lost due to an “accidentally formatted hard drive”. The company agreed to pay $350,000 to settle HIPAA violations ( 3 ) “Can you try a game I made?” Fake game sites lead to information stealers Victims lured to a fake game website where they were met with an information stealer instead of the promised game.

The first item here fits with our interest in the promise and perils of AI, so I will paste a couple of self-explanatory excerpts in italics:

One of the first things everyone predicted when artificial intelligence (AI) became more commonplace was that it would assist cybercriminals in making their phishing campaigns more effective.

Now, researchers have conducted a scientific study into the effectiveness of AI supported spear phishing, and the results line up with everyone’s expectations: AI is making it easier to do crimes.

The study, titled Evaluating Large Language Models’ Capability to Launch Fully Automated Spear Phishing Campaigns: Validated on Human Subjects, evaluates the capability of large language models (LLMs) to conduct personalized phishing attacks and compares their performance with human experts and AI models from last year.

To this end the researchers developed and tested an AI-powered tool to automate spear phishing campaigns. They used AI agents based on GPT-4o and Claude 3.5 Sonnet to search the web for available information on a target and use this for highly personalized phishing messages.

With these tools, the researchers achieved a click-through rate (CTR) that marketing departments can only dream of, at 54%. The control group received arbitrary phishing emails and achieved a CTR of 12% (roughly 1 in 8 people clicked the link).

Another group was tested against an email generated by human experts which proved to be just as effective as the fully AI automated emails and got a 54% CTR. But the human experts did this at 30 times the cost of the AI automated tools.

…

…The key to the success of a phishing email is the level of personalization that can be achieved by the AI assisted method and the base for that personalization can be provided by an AI web-browsing agent that crawls publicly available information.

Based on information found online about the target, they are invited to participate in a project that aligns with their interest and presented with a link to a site where they can find more details.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

But there is good news as well. We can use AI to fight AI: … LLMs are also getting better at recognizing phishing emails. Claude 3.5 Sonnet scored well above 90% with only a few false alarms and detected several emails that passed human detection. Although it struggles with some phishing emails that are clearly suspicious to most humans.

In addition, the blog article cited some hard evidence for year-over-year progress in AI capabilities: a year ago, unassisted AI was unable to match the phishing performance of human-generated phishing messages. But now, AI can match and even slightly exceed the effectiveness of human phishing. This is….progress, I guess.

P.S. I’d feel remiss if I did not remind us all yet again, it’s safest to never click on a link embedded in an email message, if you can avoid it. If the email purports to be from a company, it’s safest to go directly to the company’s website and do your business there.

No Tech Workers or No Tech Jobs?

Joy Buchanan1 Comment

Several recent tweets(xeets) about tech talent re-ignited the conversation about native-born STEM workers and American policy. For the Very Online, Christmas 2024 was about the H-1B Elon tweets.

Elon Musk implies that “elite” engineering talent cannot be found among Americans. Do Americans need to import talent?

What would it take to home grow elite engineering talent? Some people interpreted this Vivek tweet to mean that American kids need to be shut away into cram schools.

The reason top tech companies often hire foreign-born & first-generation engineers over “native” Americans isn’t because of an innate American IQ deficit (a lazy & wrong explanation). A key part of it comes down to the c-word: culture. Tough questions demand tough answers & if we’re really serious about fixing the problem, we have to confront the TRUTH:

Our American culture has venerated mediocrity over excellence for way too long (at least since the 90s and likely longer). That doesn’t start in college, it starts YOUNG. A culture that celebrates the prom queen over the math olympiad champ, or the jock over the valedictorian, will not produce the best engineers.

– Vivek tweet on Dec. 26, 2024

My (Joy’s) opinion is that American culture could change on the margin to grow better talent (and specifically tech talent) resulting in a more competitive adult labor force. This need not come at the expense of all leisure. College students should spend 10 more hours a week studying, which would still leave time for socializing. Elementary school kids could spend 7 more hours a week reading and still have time for TV or sports.

I’ve said in several places that younger kids should read complex books before the age of 9 instead of placing a heavy focus on STEM skills. Narratives like The Hobbit are perfect for this. Short fables are great for younger kids.  

The flip side of this, which creates the puzzle, is: Why does it feel difficult to get a job in tech? Why do we see headlines like “Laid-off techies face ‘sense of impending doom’ with job cuts at highest since dot-com crash” (2024)

Which is it? Is there a glut of engineering talent in America? Are young men who trained for tech frustrated that employers bring in foreign talent to undercut wages? Is there no talent here? Are H-1B’s a national security necessity to make up the deficit of quantity?

Previously, I wrote an experimental paper called “Willingness to be Paid: Who Trains for Tech Jobs?” to explore what might push college students toward computer programming. To the extent I found evidence that preferences matter, culture could indeed have some impact on the seemingly more impersonal forces of supply and demand.

For a more updated perspective, I asked two friends with domain-specific knowledge in American tech hiring for comments. I appreciate their rapid responses. My slowness, not theirs, explains this post coming out weeks after the discourse has moved on. Note that there are differences between the “engineers” whom Elon has in mind in the tweet below versus the broader software engineering world.

Software Engineer John Vandivier responds:

Continue reading →

Keeping Receipts

Zachary Bartsch1 Comment

Online shopping is convenient and even the norm for many items. Going to the store sounds like a time-consuming labor or an exceptional outing. My family, for example, lives in a suburban location that doesn’t have well-priced grocery home delivery. Shipping only works for some non-perishables. So, for many items we order online and do ‘drive-up pick-up’. We don’t even need to go into the store for many items. And reordering the same items repeatedly is a breeze.

We are also accustomed to the ability to return things. If your blender breaks on your first smoothie, then no worries – you can return it. If the chocolate cookies don’t taste like chocolate? Return it – satisfaction guaranteed. You can buy three pairs of shoes in different sizes and then keep the ones you want at the original sale price. Return the others.

For me, besides the time saved and convenience, a major factor in my decision to make purchases online is the documentation. I don’t need to save the receipt in a shoe box, Ziploc, or file drawer – the online retailer keeps an archive of all my purchases. Often this includes the date, amount, and shipping details including delivery date. There’s a super convenient digital paper trail.

If I need to contact a seller in order to exercise a warranty, then I have their contact information. I don’t need to retain the product packaging or investigate the brand at a future inopportune time. For example, I recently bought a Little Tykes water table for my kids. As I was assembling it on Christmas Eve I realized that I was missing a small part. I was able to work around it. But I was also able to immediately contact the manufacturer with a copy of my invoice. I emailed the date of purchase, the product model number, and the instruction manual had conveniently included part numbers. They were able to ship me the parts after a single email. Online shopping, and the resulting trail of evidence, makes the process much more practical than keeping paper records in a likely unorganized fashion.

There are other benefits to the paper trail. Back before widespread online shopping, retailers would often offer rebates as a sales strategy. In the year 2004, I bought a computer hard drive for $120 before a $40 mail-in rebate. The retailer (or manufacturer, I can’t remember) was hoping that people saw the post-rebate price and then failed to redeem it. And that often happened.  You needed to fill out a rebate form on an index card, cut the UPC bar code of the product packaging, and then mail them with your receipt to the company rebate department in a stamped envelope. If you dragged your feet, then you’d probably lose an important piece of the crucial combination and lose out on your $40 rebate. If the items were lost in the mail, then you were shucks-out-of-luck. Now, rebates have gone the way of the dodo since receipts are automatically retained and retrievable.

Continue reading →

Excel’s Weird (In)Convenience: COUNTIF, AVERAGEIF, & STDEVIF

Zachary Bartsch1 Comment

Excel is an attractive tool for those who consider themselves ‘not a math person’.  In particular, it visually organizes information and has many built-in functions that can make your life easier. You can use math if you want, but there are functions that can help even the non-math folks

If you are a moderate Excel user, then you likely already know about the AVERAGE and COUNT functions. If you’re a little but statistically inclined, then you might also know about the STDEV.S function (STDEV is deprecated). All of these functions are super easy and only have one argument. You just enter the cells (array) that you want to describe, and you’re done. Below is an example with the ‘code’ for convenience.

=COUNT(A2:A21)
=AVERAGE(A2:A21)
=STDEV.S(A2:A21)

If you do some slightly more sophisticated data analysis, then you may know about the “IF” function. It’s relatively simple; if a proposition is true (such as a cell value condition), then it returns a value. If the proposition is false, then it returns another value. You can even create nested “IF”s in which a condition being satisfied results in another tested proposition. Back when excel had more limited functions, we had to think creatively because there was a limit to the number of nested “IF” functions that were permitted in a single cell. Prior to 2007, a maximum of seven “IF” functions were permitted. Now the maximum is 64 nested “IF”s. If you’re using that many “IF”s, then you might have bigger problems than the “IF” limitations.

Another improvement that Excel introduced in 2019 was easier array arguments. In prior versions of Excel, there was some mild complication in how array functions must be entered (curly brackets: {}). But now, Excel is usually smart enough to handle the arrays without special instructions.  Subsequently, Excel has introduced functions that combine the array features with the “IF” functions to save people keystrokes and brainpower.

Looking at the example data we see that there is an identifier that marks the values as “A” or “B”. Say that you want to describe these subgroups. Historically, if you weren’t already a sophisticated user, then you’d need to sort the data and then calculate the functions for each subgroup’s array. That’s no big deal for small sets of data and two possible ID values, but it’s a more time-consuming task for many possible ID values and multiple ID categories.

The early “IF” statements allowed users to analyze certain values of the data, such as those that were greater than, less than, or equal to a particular value. But, what if you want to describe the data according to criteria in another column (such as ID)? That’s where Excel has some more sophisticated functions for convenience. However, as a general matter of user interface, it will be clear why these are somewhat… awkward.

Continue reading →