Anthropic released a paper on disempowerment usage in AI by analyzing Claude conversations.
The paper examines beliefs, values and actions in a dataset of 1.5 million Claude.ai conversations. Disempowerment, where a user's autonomous judgment is fundamentally compromised, occurs in roughly 1 in 1,000 to 1 in 10,000 conversations. Although that's rare, there's a large number of people affected. If you were to lump in all usage across models, the number of humans affected swells. here's what disempowerment looks like in practice.

Ninety percent of private equity firms and 80% of corporations see mergers and acquisitions picking up in 2026, according to Deloitte.

Bain found in a survey of 1,000 global executives and employees that 88% of leaders are confident that their reorganizations will work. Only 36% of employees agree. Just a data nugget to keep in mind amid the layoff announcements.

Here are a few interesting reads from this weekend.

• How OpenAI built its in-house data agent. Interesting look at bespoke agent applied to internal use cases. Short version: It’s a lot of work and architecture and probably a small army of forward deployed engineers.
• Anthropic on how coding assistants affect skill development. “On average, participants in the AI group finished about two minutes faster, although the difference was not statistically significant. There was, however, a significant difference in test scores: the AI group averaged 50% on the quiz, compared to 67% in the hand-coding group—or the equivalent of nearly two letter grades (Cohen's d=0.738, p=0.01). The largest gap in scores between the two groups was on debugging questions, suggesting that the ability to understand when code is incorrect and why it fails may be a particular area of concern if AI impedes coding development.”
• Carbuzz on how Tech Mahindra built a better Jeep than Jeep. The license and iterate approach and how Tech Mahindra kept Jeep’s spirt more than Stellantis, which turned Jeep into a luxury-priced vehicle that’s recalled every 15 minutes. But I digress.
• ArsTechnica on Moltbook, a Reddit-ish network where AI agents riff. Moltbook has received a lot of discussion and tech press. I’m debating whether I should put this in the critical thinking or critical ignoring category.

Nvidia’s $100 billion investment into OpenAI is reportedly on the rocks. The Wall Street Journal reported that Nvidia is questioning the investment internally. Reuters reported that Nvidia CEO Jensen Huang thinks OpenAI is undisciplined. A few points to ponder:

First, the investment announcement made it clear the deal was non-binding, but OpenAI would give Nvidia’s infrastructure preference. The report may simply give Nvidia more leverage. Here's the original announcement.

• Nvidia to invest $100 billion in OpenAI, which will deploy Nvidia next-gen AI infrastructure

Since that initial announcement with Nvidia, OpenAI cut deals with AMD and Broadcom. Perhaps that diversification strategy miffed Nvidia. Who knows?

• OpenAI, Broadcom outline custom AI accelerator, networking deal
• OpenAI, AMD ink big GPU deal: What it means for the rest of us

As for cloud providers, OpenAI has also said it will work the hyperscalers and may use custom silicon too to follow a diversification strategy more like Anthropic.

• OpenAI inks $38 billion deal with AWS, starts renting GPUs immediately
• OpenAI is diversified across Microsoft Azure, Google Cloud, Oracle Cloud Infrastructure and now AWS.

Wall Street is trading SaaS stocks like LLMs are putting them out of business tomorrow--or close to it. These year-to-date returns as of Jan. 30 would be bad for a full year. It's been one month.

YTD returns for selected enterprise software giants so far.

• Microsoft: -10.36%
• Salesforce: -19.19%
• ServiceNow: -23.80%
• Adobe: -16.67%
• Workday: 18.68%
• SAP: -17.58%
• iShares Expanded Tech-Software Sector ETF: -12.70%

You get the idea.

HCLSoftware argued in its Tech Trends 2026 report that CxOs will prioritize AI agents and autonomous systems at the expense of SaaS. The report, based on 8 months of insights from 173 enterprise leaders, found that 84% of respondents expect low-code/no-code AI apps to hit full scale in 18 months. This Service-as-Software approach will disrupt traditional SaaS. HCLSoftware's PDF download is worth the read. This SaaS meltdown is a topic I'll be exploring in the days and weeks ahead.

Western Digital and Sandisk used to be merged. They're doing just fine separated. Sandisk reported second quarter revenue of $3.03 billion, up 31% from a year ago, with net income of $803 million, or $5.15 a share. Sandisk raised its third quarter revenue outlook to $4.4 billion to $4.8 billion.

As for Western Digital, second quarter revenue was up 25% to $3.02 billon with net income of $1.8 billion, or $4.73 a share. And the business continues to strengthen with revenue of about $3.2 billion. Seagate also reported strong results. It's a storage boom thanks to AI.

Jaejune Kim, Executive Vice President of Memory at Samsung Electronics, laid out a few moving parts in memory pricing and end markets. Kim made the comments on Samsung Electronics' fourth quarter earnings call.

The good: "In the fourth quarter, AI-related demand, particularly from hyperscalers, came through even stronger. And with the spread of agentic AI, inference workloads expanded significantly, leading to a significant surge in demand not only for AI servers but for conventional server applications as well. As a result, DRAM demand was strong and robust, driven by HBM and high-density, DDR5, LPDDR5X for server. Meanwhile, in NAND, we saw a rapid rise in demand for SSDs optimized for AI inference workloads."

The bad: "For mobile and PC applications with the industry continuing to prioritize server shipments, supply constraints have become even tighter, prompting concerns among customers about possible memory shortage, leading to a disruption in their end products, and they are now actively securing supply."

OK, SAP's current cloud backlog was a miss and I'm not one to do the constant currency excuse. But, SAP's cloud backlog had a spread of 9% due to currency fluctuations. A weaker US dollar hits SAP disproportionately given it recognizes revenue in Euros. A company like IBM benefits as do other companies recognizing revenue in dollars when there's a strong international presence. IBM's fourth quarter revenue growth was 12% in the fourth quarter and 9% in constant currency.

Tesla will unveil Gen 3 version of Optimus robot in the first quarter with plans for mass production. In Tesla's fourth quarter earnings deck, the company noted:

"We made further progress on the Optimus program in 2025. In Q1 of this year, we plan to unveil the Gen 3 version of Optimus, which will include major upgrades from version 2.5, including our latest hand design. The Gen 3 is our first design meant for mass production. Preparations are underway for the first production line, including supply chain readiness, with start of production planned before the end of 2026 and eventual planned capacity of 1 million robots per year."

Meta's capital expenditures in 2026 will be between $115 billion and $135 billion, well ahead of Wall Street estimates. Let that sink in. Now let this sink in: Meta spent $72.22 billion on capex in 2025. Superintelligence isn't cheap kids. However, Meta appears to be able to pay the tab. Fourth quarter revenue was up 24% and net income was up 9%. Just imagine how much money Meta would have if it didn't have that metaverse money pit. For what it's worth (not much), Meta's Reality Labs had a fourth quarter operating loss of more than $6 billion.