Mar 27, 2024

The Economist: "technological breakthroughs take ages to pay off"

Previous technological breakthroughs have revolutionized what people do in offices. (example: typewriter in lates 1800s, rise of computer a century later; more recently the rise of work at home)...  Could generative AI prompt similarly profound changes?  A lesson of previous technological breakthroughs is that, economywide, they take ages to pay off.  The average worker at the average firm needs time to get used to new ways of working.  The productivity gains from the personal computer did not come until at least a decade after it became widely available.  So far there is no evidence of an AI-induced productivity surge in the economy at large.

~ "Meet your copilot: The inside view of how companies are using generative AI," The Economist, March 2, 2024



The Economist on bugs in generative AI

In November Microsoft launched a Copilot for its productivity software, such as Word and Excel.  Some early users find it surprisingly clunky and prone to crashing - not to mention cumbersome, even for people already adept at Office.  Many bosses remain leery of using generative AI for more sensitive operations until the models stop making things up.  Recently Air Canada found itself in hot water after its AI chatbot gave a purveyor incorrect information about the airline's refund policy.  That was embarrassing for the carrier, but it is easy to imagine something much worse.  Still, even the typewriter had to start somewhere.

~ "Meet your copilot: The inside view of how companies are using generative AI," The Economist, March 2, 2024



Mar 26, 2024

The Economist on the failure of Israel's invasion of Gaza

If you are a friend of Israel this is a deeply uncomfortable moment.  In October it launched a justified war of self-defence against Hamas, whose terrorists had committed atrocities that threaten the idea of Israel as a land where Jews are safe.  Today Israel has destroyed perhaps half of Hamas’s forces.  But in important ways its mission has failed. 

First, in Gaza, where its reluctance to help provide or distribute aid has led to an avoidable humanitarian catastrophe, and where the civilian toll from the war is over 20,000 and growing.  The hard-right government of Binyamin Netanyahu has rejected plans for post-war Gaza to be run by either the Palestinian Authority (pa) or an international force.  The likeliest outcome is a military reoccupation.  If you add the West Bank, Israel could permanently hold sway over 4m-5m Palestinians.

Israel has also failed at home.  The problems go deeper than Mr Netanyahu’s dire leadership.  A growing settler movement and ultra-Orthodox population have tilted politics to the right and polarised society.  Before October 7th this was visible in a struggle over judicial independence.  The war has raised the stakes, and although the hard-right parties of the coalition are excluded from the war cabinet they have compromised Israel’s national interest by using incendiary rhetoric, stoking settler violence and trying to sabotage aid and post-war planning.  Israel’s security establishment is capable and pragmatic, but no longer fully in charge. 

Israel’s final failure is clumsy diplomacy.  Fury at the war was inevitable, especially in the global south, but Israel has done a poor job of countering it.  “Lawfare”, including spurious genocide allegations, is damaging its reputation.  Young Americans sympathise with it less than their parents do.  President Joe Biden has tried to restrain Mr Netanyahu’s government by publicly embracing it, but failed.  On March 14th Chuck Schumer, Israel’s greatest ally in the Senate, decried Hamas’s atrocities but said Israel’s leader was “lost”.

It is a bleak picture that is not always acknowledged in Jerusalem or Tel Aviv. Mr Netanyahu talks of invading Rafah, Hamas’s last redoubt, while the hard right fantasises about resettling Gaza. Many mainstream Israelis are deluding themselves, too. They believe the unique threats to Israel justify its ruthlessness and that the war has helped restore deterrence. Gaza shows that if you murder Israelis, destruction beckons. Many see no partner for peace—the pa is rotten and polls say 93% of Palestinians deny Hamas’s atrocities even took place. Occupation is the least-bad option, they conclude. Israelis would prefer to be popular abroad, but condemnation and antisemitism are a small price to pay for security. As for America, it has been angry before. The relationship is not about to rupture. If Donald Trump returns he may once again give Israel a free pass. 

This seductive story is a manifesto for disaster.






Mar 25, 2024

Fred Hickey on the lack of demand for generative AI

Understand that I believe in AI.  But AI has been with us for decades - becoming a bigger part of our lives year after year.  But Generative AI (Gen AI) - which just came on the scene last year when Microsoft touted it as means to threaten Google's stranglehold on the search market (so far Microsoft has failed with that attempt - getting just 1% additional market share) is another matter.  To me (and many other observers) Gen AI is like the Metaverse or Blockchain - "all hat and no cattle" - as the saying goes.  For example, Wired magazine (not exactly a Luddite's favorite read) titled a recent story: "Get Ready for the Great AI Disappointment."  Wired: "More and more evidence will emerge that Generative AI an large language models provide false information and are prone to hallucination - where an AI simply makes stuff up, and gets it wrong."  Anticipation that there will be exponential improvements in productivity across the economy, or the much vaunted first steps towards 'artificial general intelligence,' or AGI will fare no better."  "Some people will start recognizing that it was always a pipe dream to reach anything resembling complex human cognition on the basis of predicting words."  There are many similar stories out there questioning the usefulness of Gen AI.

Microsoft has placed its AI-powered assistant called Copilot that uses ChatGPT (a Gen AI version from OpenAI) in the upper right-hand corner of my computer screen.  When it first came out, I tested it when I was researching topics.  I found its answers underwhelming.  I rarely use it anymore - just as I almost never utilized Microsoft's previous assistant "Clippy."  My experience with ChatGPT is not unusual.  Data traffic analytics firm Similarweb recently reported that ChatGPT web traffic has declined in five of the last eight months and is currently (January 2024 data0 11% lower than its peak in May 2023.  The thrill is gone - just don't tell Wall Street.

~ Fred Hickey, The High-Tech Strategist, February 27, 2024



Derek Au on AI: "training is computation heavy, inferencing is computation light"

Inferencing is computation light; training is computation heavy.  Kind of like crypto, mining takes more resources than validation. 

At a very high level, what happens during the training of a model is that you ingest and analyze high volumes of data in order to derive a mathematical representation of the correlation of all of the data.  For example, in order to train a model to identify cats, you would expose the training algorithm to hundreds of thousands of pictures of cats.  You iterate the training of the algorithm across all of the pictures, the result of which converges to a mathematical equation, which becomes the trained model.  The training process is very computationally intensive because you are computing across the entire training set, i.e. every single one of the cat pictures.  GPUs are uniquely suited for these tasks because many of those computations can be performed in parallel. 

Inferencing is simply pattern matching.  It is the process of making predictions, decisions, or conclusions based upon the data given to the trained model.  You take an unknown picture, run it through the trained model, and it returns a probability distribution on whether the picture is a cat.  Inferencing is computationally light relative to training because the computation involved is only running the single picture in question through the trained model (as opposed to deriving a mathematical model of a cat by computing though hundreds of thousands of pictures of various cats). 

Inferencing can be done on existing CPUs, which are better suited for sequential computing.  It is fine for this task because you are only processing the image in question.  Still, there are a number of startups creating inferencing chips that aim to be faster, and more energy efficient (not for climate change, but for battery life!)

Over many years, Nvidia has developed its own proprietary programming language for its GPUs, to which the developer community has become accustomed.  Nvidia has the lead and mindshare on this.  I believe this is a competitive moat.  While competitors may introduce rival GPUs, will the programming community adopt them if the rival GPUs lack the familiar programming resources that programmers are accustomed to?

~ Derek Au, technology analyst, Orange Silicon Valley, February 14, 2024



Sheldon Richman on the proposed bill to force a sale of TikTok

According to Glenn Greenwald, the bill had been floating around for a few years but had not garnered enough support to get through Congress.  That changed recently, according to Greenwald, citing articles in the Wall Street Journal, Economist, and Bari Weiss's Free Press.  Why?  As Greenwald documents, anxiety about TikTok took a quantum leap beginning on Oct. 7, 2023, the day Hamas killed and kidnapped hundreds of Israeli civilians and Israel began retaliating against the people of the Gaza Strip. 

What has this got to do with TikTok? you ask.  Good question.  Israel's defenders in the United States, such as Jonathan Greenblatt of the Anti-Defamation League, are upset that TikTok's young users are being exposed to what he calls anti-Zionist and anti-Semitic disinformation.  "It's Al Jazeera on steroids," Greenblatt said on MSNBCDuring a leaked phone call, he complained, "We have a TikTok problem," by which he means a generational problem.  Younger people -- including younger Jewish people -- are appalled at what Israel's military is doing in Gaza.  (To complicate things, it looks like TikTok and Instagram have suppressed pro-Palestinian information.) 

Would an American-owned TikTok be easier to control?  Experience says yes.  Have a look at the Twitter Files, which document how American officials, Chinese-style, pressured social media to censor or suppress dissenting views on important matters such as the COVID-19 response and the 2020 election.  A federal judge likened the government's efforts to the Ministry of Truth in Orwell's Nineteen Eighty-Four.

Do we want to become more like China?

~ Sheldon Richman, "TGIF: Leave TikTok Alone," Free Association, March 22, 2024



Tom Woods on his stingray experience

When I open my eyes in the morning I'm usually looking at a wall that contains (I just counted) 60 small pictures, each of which depicts a happy memory from the past five years or so.  

Last week was rather exhausting for me, so I'm relaxing in bed before heading for the gym in a little while, and one of those pictures caught my eye.  It shows my wife and me in the ocean at a stingray experience in the Bahamas.  I went primarily for her sake, but the experience was enjoyable enough. 

At one point everyone was given a chance to touch a stingray.  I politely declined.  Then the kind of thing I can never understand happened: The event organizer, along with the other attendees, started insisting I do it, thinking that maybe I was frightened or nervous.  Nope, just don't want to.  I have no idea what difference what I do makes to anyone else there, but man did they want me to.  They insisted again.  Again, I politely declined.  I did not make a spectacle of myself, I wasn't a drama queen about it, nothing.  I simply declined. 

For some reason this episode left my wife very impressed.  She couldn't believe I had simply stood my ground.  From then on, whenever there's been pressure on me from any group at all to do something I prefer not to do, she recalls the episode with the stingray.  Sometimes people hear about an incident like this and wonder if I'm just a contrarian by nature -- particularly given my views on politics and many other subjects.  But I'm not.  

However, doing the opposite of what's expected of me has generally served me well.

~ Tom Woods, March 25, 2024