Recent reports indicate that the parents of Sam Bankman-Fried, Joseph Bankman and Barbara Fried, have been exploring the possibility of securing a presidential pardon for their son from President Donald Trump. Sam Bankman-Fried, the former CEO of the now-defunct cryptocurrency exchange FTX, was sentenced to 25 years in prison for fraud. His parents, both Stanford Law School professors, have engaged in discussions with lawyers and individuals close to Trump’s circle, although it’s unclear if they have made direct outreach to the White House.

This situation comes at a time when Trump has previously shown a willingness to use his pardon power, notably with the commutation of Silk Road founder Ross Ulbricht’s life sentence. However, the context for Bankman-Fried’s case seems different, as he lacks the widespread advocacy support that Ulbricht had from the crypto and libertarian communities. The attempt to secure a pardon has sparked various reactions, with some in the crypto community and on social media expressing skepticism or opposition to such a move, given Bankman-Fried’s prior political donations to Democrats and the nature of his conviction.

The legal and political implications of this move are significant, reflecting the complexities of high-profile white-collar crime cases intersecting with presidential clemency decisions. The outcome of these efforts remains uncertain, with no confirmation from Trump or his administration on considering such a pardon.

Trump granted 237 acts of clemency during his first term, consisting of 143 pardons and 94 commutations. This number is lower compared to many modern presidents, with only George H.W. Bush and George W. Bush granting fewer over their full terms. Trump often bypassed the traditional process via the Office of the Pardon Attorney (OPA), instead relying on an ad hoc process influenced by personal and political connections. Only 25 of his clemency grants followed the OPA’s process, with the rest coming through direct appeals or connections to Trump allies.

Roger Stone, a political consultant convicted of lying to Congress, was among those whose sentences were commuted. Steve Bannon, Trump’s former chief strategist, was pardoned for charges related to fraud in a border wall fundraising campaign. Alice Marie Johnson, whose sentence for drug trafficking was commuted following advocacy by Kim Kardashian, was later fully pardoned.

Charles Kushner, father of Trump’s son-in-law Jared Kushner, received a pardon for convictions including tax evasion and witness tampering.
Other notable recipients included individuals like Paul Manafort, Michael Flynn, and George Papadopoulos, all connected to the Mueller investigation into Russian interference in the 2016 election.

Ross Ulbricht was convicted for creating and operating Silk Road, an online black market where users could buy and sell illegal goods, including drugs, using bitcoin. His crimes were primarily centered around conspiracy to commit money laundering, computer hacking, and narcotics trafficking. Ulbricht’s platform facilitated over $200 million in illegal transactions, but he himself did not steal money directly from users; rather, he enabled a marketplace for illegal activities.

Sam Bankman-Fried was convicted of fraud, conspiracy, and money laundering related to the collapse of FTX, a major cryptocurrency exchange. His charges stem from directly defrauding customers and investors by misusing billions of dollars of customer funds for personal gain and to cover losses at Alameda Research, his trading firm. Unlike Ulbricht, Bankman-Fried’s actions resulted in direct financial loss to his customers and investors, leading to the platform’s insolvency and one of the largest financial frauds in crypto history.

Over the years, Ulbricht gained significant support from parts of the crypto community, particularly those with libertarian views, who argued that his sentence was excessively harsh and that he was being punished for pioneering a new form of commerce. The #FreeRoss movement had substantial backing, which played a role in Trump’s decision to commute his sentence.

Bankman-Fried’s case has not seen the same level of public support for several reasons. His actions were seen as a direct betrayal of trust by users of FTX, leading to widespread financial loss. Additionally, his political activities and donations have been scrutinized, potentially affecting public and political sympathy. The crypto community’s reaction has been mixed, with some highlighting the disparity in sentencing but also pointing out the severity of his crimes compared to Ulbricht’s.

Microsoft has thrown a major disruptive force into the AI market, announcing free, unlimited access to OpenAI’s powerful o1 model through Copilot’s new “Think Deeper” feature.

The move undercuts OpenAI’s own pricing model, which charges users $200 per month for unlimited access via ChatGPT Pro and $20 per month for limited access via ChatGPT Plus.

By offering OpenAI’s most advanced AI reasoning model at no cost, Microsoft is positioning Copilot as an attractive, free alternative to OpenAI’s subscription-based service. The announcement, made by Microsoft AI chief Mustafa Suleyman, confirms that Copilot users “everywhere” can access o1 at no cost, provided they sign in with a Microsoft account.

This aggressive strategy underscores Microsoft’s deep integration with OpenAI, as it continues to leverage its multi-billion-dollar investment in the AI company. Unlike OpenAI’s direct-to-consumer pricing model, Microsoft is absorbing the cost and embedding AI functionality within its ecosystem, making Copilot a more appealing choice for everyday users.

The Think Deeper feature, designed to deliver more thoughtful and well-researched responses, contrasts with Copilot’s usual, more concise replies. While not a search engine, Think Deeper specializes in complex problem-solving, coding assistance, and analytical reasoning—areas where deep AI processing is critical.

This development, however, is not happening in a vacuum, it comes amid an escalating price war in the AI industry, driven by increasing competition and a recent breakthrough in AI efficiency that threatens to collapse prices across the board.

Microsoft’s move comes just days after the emergence of DeepSeek, a Chinese AI model that dramatically cut the cost of running large-scale AI models. DeepSeek’s breakthrough—a cutting-edge Mixture of Experts (MoE) approach—has allowed it to perform better than OpenAI’s GPT-4 while slashing operational costs.

DeepSeek’s innovation is shaking up the AI industry by showing that state-of-the-art AI doesn’t have to be expensive, challenging OpenAI’s high subscription fees. In response, rival AI developers, including Microsoft, are scrambling to adjust their pricing models to avoid losing market share to newer, more cost-efficient models.

Microsoft is showing that it is willing to absorb AI costs to remain competitive by offering o1 for free. This move pressures OpenAI, Google, and other AI players to either lower their own prices or risk losing users to Copilot.

Unlike OpenAI, which must monetize its AI services directly, Microsoft can afford to subsidize AI costs because it integrates AI across its massive software ecosystem—including Windows, Office, and Azure cloud services.

With the Think Deeper feature, Microsoft is providing users with an AI assistant that can generate complex reasoning, write code, and analyze intricate topics—all for free. While the company has not yet signaled plans to charge for Think Deeper, the move suggests that Microsoft may be setting the stage for a long-term AI platform play, where it dominates AI-powered search, productivity, and cloud computing.

However, Microsoft’s decision to give away access to o1 raises a critical question for OpenAI: Can it continue charging premium rates for AI models when competitors are offering similar or better technology at significantly lower costs—or even for free?

Adding pressure to OpenAI’s situation, the company has already moved forward with o3, an even more advanced model that introduces “private chain of thought” reasoning, which enhances logical processing and coding abilities. However, o3 is almost certain to remain behind a paywall, meaning OpenAI must now justify its pricing in a rapidly commoditizing AI market.

With DeepSeek lowering AI costs and Microsoft offering OpenAI’s own model for free, the AI landscape is shifting fast. The next few months could force OpenAI, Google, and other AI players to either radically rethink their pricing models or risk losing users to free alternatives like Microsoft’s Copilot.

For now, Microsoft has made its move, and it’s clear: the AI price war has begun.

Which general studies course would you recommend for all students besides the current ones in Nigerian universities? We suggest “Prompt Engineering Techniques”. You can go with “AI Prompting Essentials” or Prompting Techniques & Methods to overcome Nigerian Society of Engineers’s reservation of the word “engineering” in non-engineering programs. Let me explain why we think that way.

In most universities in Nigeria, there are courses which fall under the umbrella of General Studies (GST) and in most universities they include Use of English, Logic and Philosophy (Humanities), Polity and Economy of Nigeria, etc, understanding that many universities call them different names. Largely, these courses are required early year courses irrespective of the program the student is enrolled to undertake.

But over the last two years, we have noted that a student’s ability to ask the right questions to available technologies, including AI, is even more important than the student knowing the answer. Simply, we need to educate and train students with basic foundational means to ask the right questions and get the right answers from the machines and computing systems of the future. In that AI future, understanding the prompting technique and methodology will be useful. And I think it is as basic as the requirement of the Use of English which enables us to communicate with humans. But here, we need to communicate with machines; feel free to call it evolved Use of English for AI!

“Prompt engineering” in AI refers to the practice of carefully crafting specific instructions, called “prompts,” to guide a generative AI model to produce accurate and relevant responses by providing the right context and information for the desired task, essentially maximizing the model’s capabilities through well-designed input phrasing. This course will equip students with the right prompting skills for foundation, LLMs and GenAI levels of AI. I think every great school should make it a GST course.

In the next Tekedia Mini-MBA which begins on Feb 10, Tekedia Institute is introducing Prompting Essentials to provide the basis for our Learners to progress on this, focusing on the business management and leadership aspect; nothing technical as we’re a business school. We will demonstrate with ChatGPT, Microsoft Copilot and Google Gemini how to win the communication of the future.

Comment on Feed

Comment: But prompting is quite easy. I don’t think that you will have enough to fill a semester coursework

My Response: In Use of English, you need two semesters to learn how to communicate with humans after spending all the time in secondary school learning English. Then, the Use of Languages for Machines (yes, that is what that AI prompting is about), you think it is not a full semester course. Provided the course is not about “define prompt engineering”, “name the 3 people who gave a speech about prompt engineering you kno”, list “3 types of prompt workers in Chicago” … you get the idea of our normal Nigerian way of teaching, a semester may not even be enough.