Should we be wary of geeks bearing gifts? Or is it time to embrace AI as part of the family?
Strategic Agenda
Specialist design, editing and translation services for the UN and the international development sector.
Written by Fola Yahaya
Thought of the week: Should we be wary of geeks bearing gifts?
Last week, OpenAI kicked off a clever '12 Days of Christmas' product release campaign, starting with modest tweaks to its 'thinking' line of AI models and the release of a $200 pro version, which includes the release of OpenAI’s long-teased Sora text-to-image generator. Ironically, this coincides with a growing chorus of AI sceptics lamenting what they see as the plateauing of most AI advancements and signs of declining adoption. Critics highlight the chasm between the hype surrounding AGI (Artificial General Intelligence) and the reality: trillions of dollars spent to marginally improve tasks we already know how to do – like writing thank you letters or making homework cheating frictionless. They also call out the environmental toll, reminding us that a single AI chat session can 'cost' the planet several litres of water.
I get the frustration. But – and it’s a massive but – like ungrateful children dismissing presents under the Christmas tree, we often forget how profoundly AI has embedded itself into our lives in under two years. Many an office worker would shudder at the thought of losing their trusty, ever-willing AI assistant. And if – God forbid – our AI chat histories ever became more public than they already are, I suspect we’d all be a little embarrassed at how lazy we’ve become. Pulitzer-winning journalists and frazzled parents alike would be exposed for their increasing AI dependence to get them through the tedium of daily life.
So, let’s celebrate the remarkable strides made by AI in just two years. Let’s toast the ingenuity of OpenAI and its peers, and embrace the cognitive lift their tools provide. But we should also be mindful of one of the foundational lessons of economics: that there’s no such thing as a free lunch. Having got many of us hooked on the ease with which they can do some of our work for us, 2025 will be the year that AI companies slowly ratchet up the price of the 'pro' versions of their tools. This will be accompanied by the emergence of a growing divide between the AI 'haves' and 'have nots'.
Moreover, as we race towards AGI – defined loosely as the ability for machines to perform any intellectual task a human can – there will be a reckoning. The AI Trojan Horse has already permeated our work and lives. So enjoy creating videos with Sora, continue to get ChatGPT to do your 3T (tedious, time-consuming and tiring) tasks and marvel as AI agents fill in your forms. But just as the Greeks defeated the Trojans with their infamous wooden gift, we should be wary of geeks bearing AI gifts.
Should firms stop hiring humans?
In San Francisco, an AI start-up called Artisan has spent an untold sum blitzing the city with an advertising campaign that questions the need for human employees. The company makes what it calls ‘artisans’, or AI agents, that it claims can fully replace humans. The brilliant (in terms of garnering oodles of free press) campaign, with such provocative slogans as: “Artisans won’t complain about work-life balance” and “Artisan’s Zoom cameras will never ‘not be working’ today” has provoked an all too predictable backlash in Silicon Valley. Yet with ‘agentic’ AI seemingly on the horizon, we need to take this hype a little more seriously.
Marc Benioff, CEO of Salesforce, has spent much of 2024 boasting about how his company plans to be the largest supplier of what he calls "digital labour". Many companies are therefore planning for the unleashing of potentially billions of agents, bespoke AI systems that are trained to do a specific task and that interact autonomously with other agents. These agents will initially make workers more efficient and then eventually redundant to increase the return to capital (for shareholders). Add to this the apparent stagnation in worker productivity and, in the UK at least, a skills gap and recent government budget that has made hiring staff a lot more expensive, and it’s unsurprising that businesses are increasingly looking at immature AI systems as an option to plug skills and financing gaps; especially with a potential recession looming.
In my own business, we’ve now got a vacancy for someone to manage bids for public tenders. As Strategic Agenda works almost exclusively for international NGOs, we have to bid for tenders – an arcane, paper-based process that has been barely digitised. Anyone can bid for these tenders, though in the past, the complexity of crafting a strong submission would weed out the jokers. (See War Dogs in which Jonah Hill plays a budding bedroom arms dealer who manages to win a $300 million tender to supply Afghan forces.)
But now, with tools like ChatGPT, anyone can upload a tender document and generate a bid that’s polished enough to pass muster. What was once a nuanced mix of art and science has been reduced to glorified admin work. On the surface, this is a win for efficiency: AI rescues both tender issuers and bid writers from the drudgery of writing lengthy requirements docs and even more verbose proposals. But clearly it also undermines the integrity of the entire public procurement process.
Instead of AI exposing the flaws of relying on human-generated answers and the lowest bid, I suspect the opposite will happen. Procurement officers will feel compelled to use AI to qualify bids, creating a new cat-and-mouse dynamic. Just as job seekers tailor CVs to bypass AI-automated sifting systems, bidders will start gaming their submissions to ensure they’re machine-readable. It’s a self-perpetuating cycle – automating a broken process rather than fixing it.
So, what’s a company to do? For us, until the procurement process itself catches up (which I sincerely hope is soon) and we can find the right human, we’re experimenting with automating key elements of our bid management process. Next week we’ll break down our strategy and step-by-step approach to using no-code apps and open-source LLMs to achieve this.
Microsoft wants AI to be part of the family
This weekend, I spent some quality time with my son at a squash tournament, capped off by a viewing of Paddington 3. Without giving too much away, the film’s core theme is about family – what it means, and who we consider to be part of it. It’s fitting, then, to reflect on Microsoft’s vision for AI: not just as a tool, but as a companion so deeply integrated into our lives that it feels like a member of the family.
Microsoft has unveiled its next-generation AI assistant, Copilot Vision, which can 'see' everything on your screen and respond to you in real time via its Edge browser. This marks a fundamental shift in how we interact with technology. The goal is to replace traditional interfaces – pointing, clicking, typing – with a conversational, intuitive experience. Think of it as less operating a machine and more chatting with a friend who knows you inside out.
Recommended by LinkedIn
Mustafa Suleyman, the CEO of Microsoft AI, describes this new wave of technology as something profoundly personal. He envisions a world where AI companions become deeply ingrained in our lives, understanding our emotions, preferences and daily rhythms. He even goes so far as to describe them as “a new digital species.” These companions won’t just make us more productive; they’ll help us feel supported and understood.
“I think of it as outsourcing a lot of mental processing to a very reliable, highly accurate, completely interactive thought partner,” Suleyman says.
My issue with this is that it runs counter to a subtle but growing pushback against tech within the domestic space. From the Australian government banning social media use for under 16s, to the looming potential ban of TikTok in the US, Microsoft is trying to sell us something that we are increasingly uncomfortable with: too much tech mediating social relations. As happy as we are for AI to do our household chores and help our kids with their homework, do we really want it to intervene during a family row?
AI video of the week
During the lengthy gap between the initial teasing of Sora last year to its final, though limited global release, this week there have been two main developments. Firstly, competitors (mainly Chinese ones) have launched and even open-sourced Sora-like apps. Secondly, OpenAI has had time to develop a proper text-to-video platform that might just be worth the $200 per month that it costs to have full access to the platform. Sadly, it's not available in the UK but reviews point to key features that include the ability to generate videos of up to 20 secs at 1080p resolution, and in widescreen, vertical or square aspect ratios. Users can also bring their own assets to extend, remix and blend, or generate entirely new content from text.
Critically, they’ve developed new interfaces to make it easier to prompt Sora with text, images and videos, with a storyboard tool that lets users specify precise inputs for each frame. Most of the feedback has been lukewarm, especially about its ability to replicate fluid dynamics, but this is only because in the 10 months since Sora was teased, we expect so much more from AI. 2025 will see much better video generators and increasingly AI-mediated video workflow. Interesting times.
What we’re reading this week
Tools we’re playing with
That's all for this week! Subscribe for the latest innovations and developments with AI.
So you don't miss a thing, follow our TikTok, Spotify, YouTube, Instagram and X pages for more creative content and insights into our work and what we do.
In case you missed it, last week's editions of the Strategic AI newsletter: