Antonym: The End of SaaS Edition
Klarna puts a down-payment on disruption of the 21st Century's most successful business model. + + PLUS + + ChatGPT-o1 bonus content.
Dear Reader
I ain’t gonna lie – we’re tech and business heavy this week.
Instant Klarna’s gonna get ya
The most successful business model invented in the past 25 years is renting software to people. Software as a service (SaaS) is ubiquitous now, a growing market worth between two and three hundred billion US dollars a year.
Could it be about to be undone?
Consumer credit company Klarna recently announced that it will discontinue using established services like Salesforce and Workday. Why? Generative AI.
Instead, the company is investing heavily in developing in-house AI solutions to handle tasks traditionally managed by these platforms. CEO Sebastian Siemiatkowski believes that this strategy will make Klarna more agile and efficient, positioning it ahead of competitors (and halving his workforce).
This move challenges the long-held belief in the indispensability of Software as a Service (SaaS) platforms for large enterprises. By cutting ties with Salesforce and Workday, Klarna is betting on AI to fill the gaps. If successful, this could set a precedent for other companies to reduce reliance on third-party providers, potentially disrupting the SaaS market.
Not everyone is convinced. Marc Benioff, CEO of Salesforce, expressed scepticism in a recent press briefing, questioning how Klarna will manage without the robust support that established platforms provide. But then, if you had perfected the most successful business model of the 21st century – software as a service (SaaS) – you would say that wouldn't you?
Klarna ditching Salesforce is not the end of SaaS. Klarna’s move may come to nothing. It’s a weak signal about the future of tech. And yet…
Single-use software
And yet, this signal relates to another we’re seeing in the way that the use of generative AI is evolving: single-use apps or “ephemeral software”. Because it is becoming so fast and easy to write code, people are creating applications for a single use.
We first saw this after the Agency Hackers conference in July. We found that using an AI system like Claude 3.1 it was easier to create an interactive graph in Javascript than doing this in a spreadsheet. It wasn’t an app to build data visualisation, it was an app that was that visualisation alone. Useful for model a simple scenario. Need another one? Build it again.
The single-use app is a long way from something as complex as Salesforce, which whole corporations rely on for large parts of their operations. But the principle is the same as the one that could build home-grown versions of enterprise software. Larger corporations can spend tens of millions on Salesforce and similar software every year.
Before generative AI the decision of “build or buy?” was an easy one with software like this. But maybe those equations will be shifting now.
ChatGPT-o1 preview
Open AI released a new version of ChatGPT on Thursday this week (12th Sep). Not the long-awaited ChatGPT-5, but a new way of using its existing model.
Called “ChatGPT-o1”, it is also the latest contribution to the body of evidence supporting the idea that geeks should not be allowed to name products that other humans have to use.
It will not replace ChatGPT-4 or 4o. Effectively it adds a brilliant prompt engineer into the process of working with the system. The AI analyses your request, comes up with an approach or series of steps for solving it and then tries it out and critiques the results.
This takes longer than a typical query – in our testing some were as 20 seconds or longer – but it also shows you its work as it goes. Here’s a screenshot of it working through a challenge i set to build a business simulator:
It excels at complex tasks like coding and writing longer pieces of code. I’ll report back as we learn more.
At the moment ChatGPT-1o is in preview mode, which seems to mean that you can’t upload documents or images. It is nonetheless very impressive and immediately useful if you have a level of comfort or literacy working with AI.
This week I’ve been…
Watching…
Alien Romulus (in cinemas). So much fun. Aesthetic and story shape are back to early 80s basics and all the better for it. Particularly loved the South London bad-boys in space vibe. Highly recommended.
Writing…
AI literacy.
We will have a Brilliant Noise paper on AI literacy out very soon and (save the date) a webinar on October 2nd (11 am UK time, details and link to follow soon).
And there’s a sneak preview – an AI literacy briefing I put together based on an FT Business Book of the Year webinar this week, plus my notes. When I say “I” and “put together” – I mean it was one of the first experiments from using ChatGPT-o1
Also check out Here’s How Big Tech Took Over AI — Parmy Olson summarising her new book Supremacy: AI, ChatGPT, and the Race that Will Change the World
That’s all…
I hope you found something useful in this week’s Antonym. If so, stick a like on it for us –thanks!
Byeee…
Antony
BONUS CONTENT
Breaking the AI Barrier
How to Overcome Resistance and Build a Workforce Ready for the Future
As artificial intelligence (AI) becomes increasingly embedded in business, its potential to transform industries, streamline operations, and drive innovation is clearer than ever. But the path to widespread adoption is fraught with challenges. For many organisations, the most significant barrier to harnessing AI’s power isn’t technological—it’s human. Cultural resistance, fear of job displacement, and a lack of AI literacy are stalling progress in even the most forward-thinking companies.
So how can senior leaders overcome these barriers and build a workforce ready to embrace AI? Here’s what experts say about breaking through the resistance and fostering an AI-literate, future-ready organisation.
1. Addressing Cultural Resistance
Resistance to AI is one of the most common challenges businesses face. Employees who have built successful careers without AI tools may struggle to see the need for change. As Jeremy Hunt, Head of the Client Insights Unit at Schroders, explained during a recent Financial Times webinar, “There are some people that are going to naturally feel a little bit uncomfortable with using AI… they’ve been working very successfully for a number of years without it."
This discomfort is understandable. Many workers see AI as a threat to their roles or fear the technology will make their skills redundant. However, senior leaders must frame AI not as a replacement for human ingenuity but as a tool to enhance it. AI should be positioned as a resource that helps employees achieve more, not less—a perspective that starts at the top.
Solution: Leaders should actively communicate the benefits of AI and how it can empower teams rather than displace them. This means sharing success stories and case studies within the company where AI has made a significant, positive impact on productivity and job satisfaction.
Sherry Coutu, entrepreneur and board member of companies such as Pearson and Raspberry Pi, offered a key insight: “When you see something working—someone serving internally at 10 times the speed—hold up examples of that from leadership and say, ‘I was amazed at how fast you turned around that request. How did you do it?’” Celebrating these wins can ease anxieties and encourage AI adoption across the organisation.
2. Investing in Training and AI Literacy
A major barrier to AI adoption is the knowledge gap. In many cases, employees aren’t resisting AI because they dislike it—they’re resisting because they don’t understand it. Coutu pointed out the scale of the challenge: “Large organisations need to retrain tens of thousands, if not hundreds of thousands, of people to feel comfortable using it."
While AI tools like ChatGPT or Claude are designed to be intuitive, they can still feel overwhelming to employees who are unfamiliar with them. The key to overcoming this fear is education. By investing in training programs that teach employees how to use AI effectively, organisations can empower their teams to integrate AI into their daily tasks.
Solution: Implement structured training and upskilling programs that address different levels of AI literacy within your workforce. For example, Schroders introduced an internal AI tool called "Genie" and provided training for employees to understand and use it effectively. "We’re making sure that everyone across Schroders has the necessary training that they're doing at the moment," said Hunt. This approach ensures that AI is seen as an accessible tool for everyone, not just tech specialists.
3. Creating a Culture of Innovation
The cultural shift required for successful AI adoption extends beyond just training—it involves fostering an environment where experimentation with AI is encouraged. Employees need to feel comfortable exploring AI tools without fear of failure or backlash. Hunt noted that at Schroders, “We allow pretty much anyone across the organisation to do that initial innovation, but within a controlled environment.”
Creating this balance—allowing freedom to experiment while maintaining oversight—is essential for embedding AI into the fabric of an organisation. Employees should feel they have permission to test new tools and workflows, with support from leadership and without the fear of making mistakes.
Solution: Establish "safe zones" where employees can experiment with AI and innovate. These could be sandbox environments or innovation labs where AI tools can be trialled before being rolled out company-wide. Leaders must set the tone by championing innovation, rewarding creativity, and highlighting successes.
4. Normalising AI as Part of Daily Work
A key reason for resistance to AI is the perception that it represents a fundamental shift in how work is done. This leads to concerns that AI could replace jobs or disrupt established workflows. However, as Parmy Olson, tech columnist for Bloomberg, pointed out, “A lot of professionals... actually don’t want to tell other people that they’ve managed to get that work done more quickly because Claude helped me, or because ChatGPT helped me.”
Normalising the use of AI as a standard tool, just like email or spreadsheets, is crucial to overcoming this resistance. AI should be seen as an everyday aid to efficiency, not a novelty or a shortcut that undermines professional effort.
Solution: Encourage transparency about AI use by integrating AI tools into daily operations in a seamless way. Provide clear messaging that AI use is not only acceptable but expected. By setting this expectation, leaders can help destigmatise the use of AI and foster a more open attitude toward its benefits.
5. Championing Success and Celebrating Early Wins
One of the most effective ways to reduce resistance is by showcasing AI’s tangible benefits. Employees are more likely to adopt new technology if they see its positive impact on their peers. Coutu noted the importance of highlighting success stories, saying, "Celebrating where we are seeing those productivity gains… allows peers to inform others of something that is working."
This is especially important in industries where the AI-enabled productivity gains are measurable. Olson compared the current stage of AI adoption to the early days of email, when the benefits weren’t immediately quantifiable but became indispensable over time.
Solution: Leaders should make it a priority to celebrate and broadcast AI success stories, both internally and externally. Recognising and rewarding employees who excel with AI tools can create a ripple effect across the organisation, accelerating adoption and dispelling myths about AI’s limitations.
Conclusion
Overcoming the barriers to AI adoption requires leadership that is proactive, transparent, and focused on building a culture of innovation. By addressing employee fears, investing in education, and celebrating early wins, senior leaders can guide their organisations through the AI transformation. AI is not just a tool for efficiency—it’s a competitive edge that will define the future of work. For businesses that embrace it now, the rewards will be substantial.
As Hunt noted, “The real magic is going to come when organisations are able to crack the end-to-end processes or even bring new products to market.” The question for senior leaders is: Are you ready to harness that magic?
The reality is, organisations that fail to act now risk being left behind in the AI race. AI's potential to revolutionise industries—from healthcare to finance to education—means that delaying adoption could widen the gap between those who embrace AI and those who resist it. As Sherry Coutu emphasised, “The greater risk is delaying the use of it, rather than adopting it. We can iterate, we can fix it…if our businesses here are not using it, and a business in the United States is using it, who’s going to win the competitive war?”
Actionable Next Steps for Senior Leaders*
1. Communicate the Value of AI: Make it clear how AI can benefit your team. Frame it as a tool that enhances capabilities, not a replacement for human jobs.
2. Invest in Training: Equip your workforce with the skills they need to use AI confidently. Tailored training programs should target both entry-level employees and senior staff.
3. Foster a Culture of Innovation: Encourage experimentation with AI in a controlled environment. Make space for creativity and innovation by empowering teams to explore AI tools.
4. Normalise AI Usage: Make AI a routine part of your organisation’s workflow. Highlight its use as standard practice to reduce stigma or feelings of inadequacy.
5. Celebrate Success: Showcase examples of where AI has improved efficiency or delivered tangible benefits. Recognising these successes can encourage further adoption across the business.
As AI continues to reshape industries and business practices, the time to act is now. Waiting risks more than just lagging behind in the adoption curve—it means missing out on AI’s transformative potential. As Parmy Olson summed up, “We are in the foothills here,” but with the right leadership, companies can soon find themselves at the summit, leading the way in an AI-powered future.
The choice is simple: act now, or be left behind. The future won’t wait, and neither should you.