Connect with us

Published

on

See also: Parrots, paperclips, and safety vs ethics: Why the artificial intelligence debate sounds like a foreign language

Here’s a list of some terms used by AI insiders:

AGI — AGI stands for “artificial general intelligence.” As a concept, it’s used to mean a significantly more advanced AI than is currently possible, that can do most things as well or better than most humans, including improving itself.

Example: “For me, AGI is the equivalent of a median human that you could hire as a coworker, and they could say do anything you would be happy with a remote coworker doing behind a computer,” Sam Altman said at a recent Greylock VC event.

AI ethics describes the desire to prevent AI from causing immediate harm, and often focuses on questions like how AI systems collect and process data and the possibility of bias in areas like housing or employment.

AI safety describes the longer-term fear that AI will progress so suddenly that a super-intelligent AI might harm or even eliminate humanity.

Alignment is the practice of tweaking an AI model so that it produces the outputs its creators desired. In the short term, alignment refers to the practice of building software and content moderation. But it can also refer to the much larger and still theoretical task of ensuring that any AGI would be friendly towards humanity.

Example: “What these systems get aligned to — whose values, what those bounds are — that is somehow set by society as a whole, by governments. And so creating that dataset, our alignment dataset, it could be, an AI constitution, whatever it is, that has got to come very broadly from society,” Sam Altman said last week during the Senate hearing.

Emergent behavior — Emergent behavior is the technical way of saying that some AI models show abilities that weren’t initially intended. It can also describe surprising results from AI tools being deployed widely to the public.

Example: “Even as a first step, however, GPT-4 challenges a considerable number of widely held assumptions about machine intelligence, and exhibits emergent behaviors and capabilities whose sources and mechanisms are, at this moment, hard to discern precisely,” Microsoft researchers wrote in Sparks of Artificial General Intelligence.

Fast takeoff or hard takeoff — A phrase that suggests if someone succeeds at building an AGI that it will already be too late to save humanity.

Example: “AGI could happen soon or far in the future; the takeoff speed from the initial AGI to more powerful successor systems could be slow or fast,” said OpenAI CEO Sam Altman in a blog post.

Foom — Another way to say “hard takeoff.” It’s an onomatopeia, and has also been described as an acronym for “Fast Onset of Overwhelming Mastery” in several blog posts and essays.

Example: “It’s like you believe in the ridiculous hard take-off ‘foom’ scenario, which makes it sound like you have zero understanding of how everything works,” tweeted Meta AI chief Yann LeCun.

GPU — The chips used to train models and run inference, which are descendants of chips used to play advanced computer games. The most commonly used model at the moment is Nvidia’s A100.

Example: From Stability AI founder Emad Mostque:

Guardrails are software and policies that big tech companies are currently building around AI models to ensure that they don’t leak data or produce disturbing content, which is often called “going off the rails.” It can also refer to specific applications that protect the AI from going off topic, like Nvidia’s “NeMo Guardrails” product.

Example: “The moment for government to play a role has not passed us by this period of focused public attention on AI is precisely the time to define and build the right guardrails to protect people and their interests,” Christina Montgomery, the chair of IBM’s AI ethics board and VP at the company, said in Congress this week.

Inference — The act of using an AI model to make predictions or generate text, images, or other content. Inference can require a lot of computing power.

Example: “The problem with inference is if the workload spikes very rapidly, which is what happened to ChatGPT. It went to like a million users in five days. There is no way your GPU capacity can keep up with that,” Sid Sheth, founder of D-Matrix, previously told CNBC.

Large language model — A kind of AI model that underpins ChatGPT and Google’s new generative AI features. Its defining feature is that it uses terabytes of data to find the statistical relationships between words, which is how it produces text that seems like a human wrote it.

Example: “Google’s new large language model, which the company announced last week, uses almost five times as much training data as its predecessor from 2022, allowing its to perform more advanced coding, math and creative writing tasks,” CNBC reported earlier this week.

Paperclips are an important symbol for AI Safety proponents because they symbolize the chance an AGI could destroy humanity. It refers to a thought experiment published by philosopher Nick Bostrom about a “superintelligence” given the mission to make as many paperclips as possible. It decides to turn all humans, Earth, and increasing parts of the cosmos into paperclips. OpenAI’s logo is a reference to this tale.

Example: “It also seems perfectly possible to have a superintelligence whose sole goal is something completely arbitrary, such as to manufacture as many paperclips as possible, and who would resist with all its might any attempt to alter this goal,” Bostrom wrote in his thought experiment.

Singularity is an older term that’s not used often anymore, but it refers to the moment that technological change becomes self-reinforcing, or the moment of creation of an AGI. It’s a metaphor — literally, singularity refers to the point of a black hole with infinite density.

Example: “The advent of artificial general intelligence is called a singularity because it is so hard to predict what will happen after that,” Tesla CEO Elon Musk said in an interview with CNBC this week.

Continue Reading

Technology

CNBC Daily Open: November hasn’t been kind — or typical — for U.S. stocks

Published

on

By

CNBC Daily Open: November hasn't been kind — or typical — for U.S. stocks

Traders work on the floor at the New York Stock Exchange (NYSE) in New York City, U.S., Nov. 26, 2025.

Brendan McDermid | Reuters

The U.S. stock market was closed Thursday stateside for Thanksgiving Day and will reopen on Friday until 1 p.m. ET.

With approximately just 3 hours of trading left for the month, major U.S. indexes are looking to end November in the red, based on CNBC calculations.

As of Wednesday’s close, the S&P 500 was down 0.4% month to date, the Dow Jones Industrial Average 0.29% lower during the same period and the Nasdaq Composite retreating 2.15%, vastly underperforming its siblings as technology stocks stumbled in November.

Unless there’s a huge jump in stocks during the shortened trading session on Friday stateside — which might not be an unequivocally positive move since it would raise more questions about the market’s sustainability — that means the indexes are on track to snap their winning streaks. The S&P 500 and Dow Jones Industrial Average have risen in the past six months, and the Nasdaq Composite seven.

It will also mark a divergence from the historical norm. The S&P 500 has advanced an average of 1.8% in November since 1950, according to the Stock Trader’s Almanac. And in the year following a U.S. presidential election, it typically rises 1.6%.

But it’s not been a typical post-presidential election year. It’s hard to see the market, in the coming months, or even years, moving according to any historical trajectory.

What you need to know today

U.S. futures are mostly flat Thursday night. The stock market was closed during the day for Thanksgiving in the U.S. Asia-Pacific markets traded mixed Friday. Japan’s Nikkei 225 ticked up in volatile trading after Tokyo inflation came in hotter than expected.

Trump to suspend migration from ‘Third World Countries.’ The U.S. president will also cancel federal benefits and subsidies to “noncitizens” in the country, he said in Truth Social posts on Thursday night stateside. Trump did not specify which countries would be affected.

South Korea imposes sanctions on Prince Group. The Cambodian conglomerate is accused of running large-scale fraud operations across Southeast Asia. The U.S., U.K. and Singapore have also imposed punitive measures on the company.

Russia is ready for ‘serious’ discussions for peace. The U.S.-led framework “can be the basis for future agreements,” Russian President Vladimir Putin said Thursday, as translated by Reuters. He added that the U.S. seemed to take Moscow’s position “into account.”

[PRO] Bank of America doesn’t see much upside for 2026. The S&P 500 should rise by a single-digit percentage point, a slowdown from recent years because one supporting factor will be shrinking, said a strategist from the bank.

And finally…

An operator works at the data centre of French company OVHcloud in Roubaix, northern France on April 3, 2025.

Sameer Al-doumy | Afp | Getty Images

Europe’s slow and steady approach to AI could be its edge

It’s unlikely that Europe will lead in building facilities for AI hyperscalers or for the training of AI — that race is considered all but won — but the general consensus is that it could excel in smaller, cloud-focused and connectivity-style facilities.

Europe has “a lot of constraints, but, actually, the more difficult something is to replicate, the more long-term value what you’ve got has,” said Seb Dooley, senior fund manager at Principal Asset Management.

— Tasmin Lockwood

Continue Reading

Technology

Baidu is emerging as a major AI chip player in China to fill the Nvidia gap

Published

on

By

Baidu is emerging as a major AI chip player in China to fill the Nvidia gap

A general view of the Baidu logo is seen at the Shanghai New Expo Center during the World Artificial Intelligence Conference 2025 in Shanghai, China, on July 28, 2025.

Ying Tang | Nurphoto | Getty Images

Tech giant Baidu is emerging as one of China’s key artificial intelligence chip players, positioning itself as a challenger to Huawei as both look to fill the void left by industry leader Nvidia being kept out of the country.

Best-known as China’s biggest search business, Baidu has in recent years refocused its business around driverless cars and AI, including a majority-owned subsidiary, Kunlunxin, which designs chips.

Several analysts have upgraded their outlook on Baidu’s stock over the past few weeks, citing the semiconductor business and forecasting the unit will gain more domestic orders.

This month, Baidu laid out a five-year roadmap for its Kunlun AI chips, beginning with the M100 in 2026 and the M300 in 2027. The company already uses a mix of its self-developed chips in its data centers to run its ERNIE AI models, as well as Nvidia products.

Baidu makes money by selling its chips to third parties building data centers as well as renting out computing capacity via its cloud. It has sought to position itself as a so-called “full stack” AI offering with infrastructure made up of chips, servers and data centers, as well as AI models and applications.

And the chip business appears to be gaining traction. Earlier this year, Kunlunxin won orders from suppliers to China Mobile, one of the country’s biggest mobile carriers.

“Kunlunxin has emerged as a leading domestic AI chip developer, focusing on high- performance AI chips for large language model (LLM) training and inference, cloud  computing, and telecom and enterprise workloads,” analysts at Deutsche Bank said in a note this month.

While Nvidia’s graphics processing units (GPUs) are widely regarded as the most advanced chips for training and running AI, the company has been blocked by the U.S. government from selling its top-end product to China. Beijing has also reportedly been persuading local tech companies not to buy the H20, a less powerful Nvidia chip designed for the Chinese market and greenlit for export.

With Huawei — the leading player through its massive clusters of chips — out of the picture, analysts are suggesting Baidu will fill the void and its chip business is set for explosive growth.

“We believe domestic demand for AI compute in China remains intense, and hyperscalers are increasingly sourcing from local solution providers,” JPMorgan said in a note on Sunday. “We view Kunlun AI chip as one of the best positioned.”

The investment bank analysts forecast Baidu chips sales to increase six-fold to reach 8 billion Chinese yuan ($1.1 billion) in 2026.

Analysts at Macquarie estimate that Baidu’s Kunlun chip unit could be valued at about $28 billion.

Baidu is not alone among China’s tech giants when it comes to self-developed semiconductors. CNBC reported in August that Alibaba is also developing its next-generation AI chip.

AI chip shortages hit China

Baidu’s chip push comes as Chinese tech giants this month said they’re seeing supply shortages.

Eddie Wu, CEO of Alibab, said that “the supply side is going to be a relatively large bottleneck” over the next two-to-three years, referring to components and chips required to build data centers.

Tencent said this month that its 2025 capital expenditure would be lower than initially anticipated. But Tencent President Martin Lau said this this was not because of a lack of demand, but more a shortage of available chips to spend the money on.

“It is not a reflection of our change in AI strategy … It is indeed a change in terms of the AI chip availability,” Lau said.

How Alibaba quietly became a leader in AI

Part of this shortage has been driven by global demand and resulting bottlenecks in the semiconductor supply chain. But China’s effective block of Nvidia chips has also reduced the supply.

Chinese tech firms have tried to mitigate the shortage by using stockpiled chips, as well as trying to make their AI models more efficient to do more with the semiconductors they have.

Meanwhile, China has its own challenges with manufacturing because its biggest chipmaker SMIC, is unable to compete on the scale and technology with leaders like Taiwan Semiconductor Manufacturing Co. That makes it hard for the China to manufacture enough domestic chips to fill the shortfall.

Like their U.S. counterparts, Chinese tech companies have continually reported strong demand for AI.

“We see that customer demand for AI is and remains very strong. In fact, we are not even able to keep pace with the growth in customer demand … in terms of the pace at which we can deploy new servers,” Alibaba’s Wu said this week.

That gives Baidu an opportunity in China.

“Baidu’s chip push is both a necessity and an opportunity. It’s a necessity, because Chinese platforms can no longer assume a steady diet of US GPUs; opportunity, because there’s now a semi‑captive, multi‑billion‑dollar domestic market for AI hardware that is compliant with both US export rules and Beijing’s self‑reliance agenda,” Nick Patience, practice lead for AI at The Futurum Group, told CNBC.

“If Baidu can ship competitive Kunlun generations on time, it doesn’t just solve its own supply problem — it becomes a strategic supplier to the rest of China’s AI industry.”

Continue Reading

Technology

CNBC Daily Open: A rough and historically atypical November for U.S. stocks

Published

on

By

CNBC Daily Open: A rough and historically atypical November for U.S. stocks

Traders work on the floor at the New York Stock Exchange (NYSE) in New York City, U.S., Nov. 26, 2025.

Brendan McDermid | Reuters

The U.S. stock market was closed Thursday stateside for Thanksgiving Day and will reopen on Friday until 1 p.m. ET.

With approximately just 3 hours of trading left for the month, major U.S. indexes are looking to end November in the red, based on CNBC calculations.

As of Wednesday’s close, the S&P 500 was down 0.4% month to date, the Dow Jones Industrial Average 0.29% lower during the same period and the Nasdaq Composite retreating 2.15%, vastly underperforming its siblings as technology stocks stumbled in November.

Unless there’s a huge jump in stocks during the shortened trading session on Friday stateside — which might not be an unequivocally positive move since it would raise more questions about the market’s sustainability — that means the indexes are on track to snap their winning streaks. The S&P 500 and Dow Jones Industrial Average have risen in the past six months, and the Nasdaq Composite seven.

It will also mark a divergence from the historical norm. The S&P 500 has advanced an average of 1.8% in November since 1950, according to the Stock Trader’s Almanac. And in the year following a U.S. presidential election, it typically rises 1.6%.

But it’s not been a typical post-presidential election year. It’s hard to see the market, in the coming months, or even years, moving according to any historical trajectory.

What you need to know today

U.S. futures are mostly flat Thursday night. The stock market was closed during the day for the Thanksgiving break in the U.S. Europe’s Stoxx 600 inched up 0.14%, rebounding from earlier losses.

Alibaba’s AI glasses go on sale. The Quark AI Glasses come in two variants that cost 1,899 Chinese yuan ($268) and 3,799 yuan, less than Meta’s $799 Meta Ray-Ban Display glasses, signaling Alibaba’s competitive entry into the consumer AI market.

Apple files a case against India’s antitrust body. The Competition Commission of India is investigating complaints about Apple’s in-app purchase policies, and could fine the company based on its global turnover — which means a potential $38 billion penalty.

Russia is ready for ‘serious’ discussions for peace. The U.S.-led framework “can be the basis for future agreements,” Russian President Vladimir Putin said Thursday, as translated by Reuters. He added that the U.S. seemed to take Moscow’s position “into account.”

[PRO] Bank of America doesn’t see much upside for 2026. The S&P 500 should rise by a single-digit percentage point, a slowdown from recent years because one supporting factor will be shrinking, said a strategist from the bank.

And finally…

An operator works at the data centre of French company OVHcloud in Roubaix, northern France on April 3, 2025.

Sameer Al-doumy | Afp | Getty Images

Europe’s slow and steady approach to AI could be its edge

It’s unlikely that Europe will lead in building facilities for AI hyperscalers or for the training of AI — that race is considered all but won — but the general consensus is that it could excel in smaller, cloud-focused and connectivity-style facilities.

Europe has “a lot of constraints, but, actually, the more difficult something is to replicate, the more long-term value what you’ve got has,” said Seb Dooley, senior fund manager at Principal Asset Management.

— Tasmin Lockwood

Continue Reading

Trending