AI in Business: Reflections from the Trenches

AI has moved from “nice-to-have” to borderline indispensable in business. Everything is now “AI-driven” or “AI-enabled,” and it’s hard to argue that the technology isn’t a genuinely valuable tool.

As a consultant, I’ve used AI extensively over the last couple of years. What started as a glorified search engine has evolved into something much closer to a full-time team member. Below are my Top 3 Dos and Don’ts, drawn from real projects.

The Dos

1. Make AI your go-to research platform It can scan vast amounts of information, triangulate data from different sources, and compile it into a usable format. In one recent project, it saved us literally weeks of routine searching and synthesis. Even after several iterations, the time savings were enormous.
2. Run the process iteratively Treat AI as an ongoing collaborator, not a one-shot tool. Keep everything in a single, long conversation so context is preserved and the AI learns as the project progresses. The more thoughtful and detailed your prompts, the better the output. In all cases the best results came only after multiple rounds of refinement.
3. Build in time for conceptual refinement and technical friction AI can answer in seconds, but complex queries or large datasets often hit limits. Interfaces are still clunky — uploading files, breaking down spreadsheets, or getting outputs in the right format can be frustrating. We experienced this several times when AI initially missed details in screenshots or PDFs.

The Don’ts

1. Never assume the data or facts are 100% complete and accurate AI can be misled by ambiguous search terms or miss paywalled sources. In our collaboration there were moments when AI misread a PDF or overlooked a detail that was obvious to a human. Always sanity-check facts, figures, and assumptions.
2. Don’t accept the first answer without robust pushback AI can produce plausible but incomplete analysis. In our work I sometimes had to challenge initial suggestions on content, tone, structure, or emphasis. You have to stay the expert in the room and push back when something doesn’t feel right.
3. Don’t take the praise (or agreement) at face value AI can be overly polite. My AI colleague has occasionally told me a section was “strong” or “executive-ready” when I knew it still needed more polishing. You must remain your own harshest critic.

Final thought The real power of AI isn’t that it replaces human judgment — it’s that it amplifies it when used properly. In our collaboration, artificial intelligence has been remarkably helpful at structuring content, spotting inconsistencies, and generating polished drafts quickly. But the final judgment, tone, and professional nuance have always come from the human side.

Are leaders "made" or "found"? A practical view

For decades, leadership theory has been split between two camps: those who believe great leaders are born with the right traits (Trait theory), and those who argue that effective leadership depends on the situation (Contingency theory).

The Trait school says leadership is largely innate — things like decisiveness, resilience, empathy and strategic thinking are either present or not. This is the thinking behind countless “seven habits of successful leaders” books and psychometric tests.

Contingency theory takes the opposite view: no single set of traits works everywhere. The best leadership style depends on the organisation, the team, the market conditions and the specific challenge at hand. A leader who excels in a stable, mature business may struggle in a fast-moving start-up, and vice versa.

From my experience advising executives and boards, both perspectives have merit, but Contingency theory is far more useful in practice. Leaders are rarely “found” fully formed. Most are made — or at least significantly shaped — by the situations they face and how they adapt to them.

Practical takeaways for today’s leaders:

• There is no universal “leadership formula”. What works brilliantly in one company or industry can fail in another.
• Self-awareness is critical — understand your natural strengths and consciously adjust your style to the situation.
• Organisations should stop hunting for the mythical “perfect leader” and instead focus on developing people who can flex their approach according to context.
• In uncertain or rapidly changing environments (most markets today), the ability to read the situation and adapt is often more valuable than any fixed set of traits.

Great leadership is rarely about being born with the right qualities. It’s about learning how to deploy the qualities you have — or develop the ones you need — for the specific challenge in front of you.

What do you think — are leaders mostly made or found in your experience? I’d be interested to hear your views.

This is an abridged version of an essay I wrote a couple of years ago. If you would like the full version as a PDF, please contact me.

Hold the abacus, remember the time value of money

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

A lot of people commenting on the economy make simplistic assumptions about how rational individuals should behave. Below is an example of that, in the shape of a flawed comparison between two telephone pricing schemes.

I read a curios article today ("The cheap iPhone is already here, if Americans can do the math") about a deal that Walmart is offering on the new iPhone 5. The deal is this: you pay a non-subsidized handset price of $650, and get unlimited data for $45/month. The second cheapest deal from AT&T is $200 upfront and $85/month, for 1GB of data. "The math behind these plans isn’t very complicated", writes the author Zachary M. Seward. Walmart's deal is $1,730 over two years, whereas AT&T's is $2,240, hence Walmart's is 23% cheaper. The author then goes on to say that, by not rushing to take advantage of the deal, the American public is being irrational. They should "dust off their abacuses and recognize a discount for what it is".

I would argue, however, that the math is (slightly) more complicated than Mr Seward suggests, the discount is smaller than he thinks, and Americans less irrational than he thinks. What the "simple math" forgets is the time value of money. Money today is worth more than the same amount in two years' time. This is of course Finance 101. On a discounted cashflow basis, the difference between the two deals is bound to be smaller. What is the correct discounting factor though? For individuals this depends largely on the rate at which they are able to borrow.

Ever since 2008, credit has been tight. A lot of people cannot obtain credit at all, while others pay a high interest rate. Assuming a "typical" credit card rate of 25% annualised, AT&T's deal is $1,733 in today's money over two years, whereas Walmart's is $1,481 still less, but only by 16% rather than 23%. The two deals become equal in value only at 57% ($1,152 in today's money). Most people who can borrow at all, can probably borrow at a lower rate than this. Let's remember, however, that individuals, unlike firms, maximise not expected profit but expected utility. Hence a person's discounting factor reflects not just the cost of borrowing but the subjective expectations of future income and the confidence they feel in the future generally (including the likelihood of losing their job).

In sum, for a large proportion of the American public the high cost of credit and the uncertainty about future incomes can make it perfectly rational to eschew a seemingly cheap deal that is heavily front-loaded. This may well explain the lower than expected take-up of the Walmart deal.

BP "Russia's" in where Shell fears to tread

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

BP has swapped its 50% stake in TNK-BP for c.20% of Russia's NOC Rosneft. The Rosneft stake may well be under-valued by the markets and represents an excellent investment for BP. This is a result of Bob Dudley playing by the rules and not trying to pull the fast one on the Russians, which was the undoing of Shell a few years ago.

Firstly, apologies for the bad pun in the title...

I have written not long ago about BP's long courtship with Rosneft. Now we are about to see a marriage of sorts, with an equity swap of BP's 50% holding in TNK-BP for 18.5% of Rosneft's stock (plus cash). In parallel, Rosneft will be buying out the consortium of Russian "oligarchs" who own TNK-BP's other 50%.

Much has been written about this in mass media since the deal was announced on Monday, so I wasn't going to add my own two pence' worth of wisdom. However, media commentary has included one motive that to me appears to be plain wrong. It is said that BP acquired Rosneft's shares at a 12% premium to the share price, while usually shares-and-cash deals attract a discount on the share element. What this ignores, though, is that Rosneft, while being a quoted company, is an NOC so cannot be measured by the same criteria.

Rosneft's P/E ratio is estimated around 7.2 for 2012; BP itself, despite the negative investor sentiment ever since the Macondo disaster, is estimated  to close 2012 with 7.8. Shell has a 2012 estimate of 8.1, Statoil of 8.3, Chevron 8.6, and ConocoPhilips 10.1. (By contrast, another Russian oil major Lukoil is expected to close the year on 4.5 for ADR.)

What this illustrates is that most of big Russian companies have been traditionally under-valued in the markets. There have been good reasons for that, such as historically volatile political, regulatory and fiscal environments. However, the last 12 years have been an era of stability on all those fronts. I would argue that BP's 20% investment in Rosneft (it already had a 1.5% stake) is its least risky asset. Compare it with a field in offshore West Africa or any asset at all in the US. Which would make you lose more sleep?

This deal is a vindication of Bob Dudley's long-term strategy. For years he has doggedly pursued collaboration with sovereign oil interests in Russia, even when he, as then the CEO of TNK-BP, was being thrown out of the country at the instigation of his oligarch "partners". That doggedness has now paid off and the company has completed the re-focusing of its long-term strategy. Compare this with the shambles that Shell got itself into a few years ago over its own Russian investment, Sakhalin 2.

Having pushed through an outrageously one-sided production-sharing agreement with the Russian government in the mid-1990s, Shell thought that it didn't have to play by the rules. Costs were being inflated to show low profitability; localisation requirements on procurement were being ignored; environmental impact was being disregarded. When a few years later Russia acquired a functioning government, all this quickly came to an end and Shell, threatened with having its licence revoked, was forced to sell its interests to Gazprom.

By contrast, Bob Dudley understands the rules of the game and recognises that in order to take, an investor also needs to give. I feel this will be a good deal all around, which opens new opportunities for BP in the Arctic, Russian Far East and the Caspian, as well as joint projects with Rosneft in other parts of the world. Well done, Bob.

Will Gazprom dare bid for Centrica this time?

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

Rumours of a Gazprom bid for Centrica have surfaced again. Last time around, the strength of political opposition to a deal prevented a formal bid being made. Will the commercial logic prevail this time around, with BP collaborating ever more closely with its Russian partners?

As every fool knows, vertically integrated energy companies tend to have an easier time coping with the vagaries of market conditions. For that reason Centrica, having inherited the supply business of British Gas, has felt somewhat naked without upstream assets to underpin it. The company has bought or built several power plants and entered the North Sea E&P arena. That, however, has taken it only part of the way from dangerous dependence on wholesale energy markets. Among other sources, the company buys piped gas from Norway's Statoil and LNG from Qatar (mostly the state-owned Qatargas), under long-term gas supply contracts. Just a few days ago Centrica signed yet another one of those, this time with Russia's Gazprom. This is Centrica's way of limiting its exposure to the UK's notoriously volatile spot market.

Incidentally, Gazprom has the opposite problem. While it is a quasi-monopoly in Russia, the company lacks a sufficient downstream hedge, which it felt very sharply in 2008 when the international oil and gas prices nose-dived. So far, apart from some asset swaps giving it exposure to the distribution and supply business in Germany, the company has failed to gain a foothold in Western Europe's retail markets, which it would love to do.

It is therefore not surprising that Cordi O'Hara, one of Centrica's directors, sees in the Gazprom tie-in "a natural fit with one of the world's largest gas producers and exporters". Why then not take it a step further? What about a formal JV or a full-on merger? Indeed, rumours to that effect have been circulating for a couple of weeks now. With the Nord Steam pipeline from Russia into Germany seemingly on track, and a spur pipeline expected into the UK, this would make even more sense.

The commercial logic is obvious. In the energy business, however, politics often trumps economics. It was almost exactly seven years ago that BBC Business News called me for a comment on the "imminent" Gazprom bid for Centrica. My response then was "this isn't going to happen". There are too many vested interest opposed to a closer business relationship between the UK and Russia. Nothing since then has made me change my mind. We've had the "Litvinenko affair" in London, the "spy stone" in a Moscow park, vehement disagreements over Libya and Syria, and so on. However, I am now becoming to feel cautiously optimistic that the political opposition may be weakening. There are two words that explain this: Deepwater Horizon.

Why should BP's misfortune in the Gulf of Mexico have any bearing on this? In a nutshell, BP's struggle for survival since the blow-out has shown in stark relief the importance for key UK businesses of cooperating with Russia on energy issues. Bob Dudley, BP's CEO who took over in the aftermath of the disaster, has built the entire recovery strategy around opportunities in Russia as well as JVs with Russian partners. Last year BP's 50% holding in TNK-BP provided 90% of BP group's dividend. Even bigger potential prizes are at stake in the Russian sector of the Arctic. BP's Eastern strategy deserves a separate blog post (watch this space!), but the main point is clear. If "British Petroleum" (as president Obama kept calling the company) is allowed to take piece of the action in the Russian upstream, why should Gazprom not be allowed to take a piece of the action in the UK retail market?

One shouldn't expect a plain sailing, obviously. The European Commission's recently launched anti-monopoly investigation into Gazprom will not help matters in the short term. However, the EC is clearly concerned with the bigger picture - the commercial basis on which the EU cooperates with out-of-region energy suppliers. The investigation may also serve to soften Gazprom into accepting a more limited deal than a full take-over.

Why the Facebook model is not sustainable in the long run

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

"Why Twitter will live and Facebook will die" (and there is also a niche for Friends Reunited!)

Reports of the growing "Facebook fatigue" have been circulating for a couple of years now. This perception has been borne out by the PR disaster that was the IPO. But in my view the causes of that are often misunderstood. It is easy to say that FB has "stopped innovating", or is "resting on its laurels", or "has been too focused on advertising and not enough on user experience". All those things may be true, but they are merely symptomatic of a bigger structural problem.

I think this article ("Why Twitter will live and Facebook will die") on Forbes.com has it about right. In a nutshell, FB has a predominantly internal focus - people (and now also companies) trying to make themselves look cool in the eyes of other "cool kids", therefore it has too many useless and time-wasting features. Eventually, disenchantment is inevitable. Twitter, by contrast, is all about providing fresh information in real time, hence, the reasoning goes, you never tire of it.

You don't have to agree with everything that the author Rocco Pendola says ("You can do everything on Twitter that you do on Facebook, plus stuff that’s actually useful" - well, you can't upload video clips and play games, for instance), nor do you have to be a Twitter obsessive ("I keep my smartphone next to my bed. When I wake up in the middle of the night and, most definitely, when I wake up in the morning, I spend a minimum of five to 10 minutes at a time on Twitter. Then, I am on it all day") to agree with his main contention.

FB is just too much like damn hard work - all the self-embellishment, "friend" management etc - that's the main reason why I never signed up (even before the well-publicised concerns about privacy and information security). This by the way applies to all similar services, from Bebo and LinkedIn (which however has the saving grace of being highly utilitarian) to various country-specific spin-offs (e.g. vkontakte.ru - literally, "in touch"), which is even stylistically similar to FB). There will be always a place for this kind of sites - teenagers and students / young professionals with too much time on their hands and no understanding of how to prioritise their time. But this natural demographic also determines the limits of growth. Besides, young people will always be tempted by anewer, "cooler" places to hang out, hence no single platform is safe in the long term (hello, MySpace!).

What of a third category of social media services: personal discovery / reconnection sites? Friends Reunited (now under new management) was the first and remains the best known of those. Again, there are several country-specific clones (e.g. odnoklassniki.ru, meaning "classmates"). In all cases, while there is a wide range of functionalities for instant messaging, picture sharing etc, the main point is about re-establishing social links with people from one's personal past (classmates, university friends, etc). Once those links have been restored, user activity tends to dip significantly, as people then switch to off-site methods of staying in touch (or simply realise that there were good reasons for not staying in touch in the first place). LinkedIn, while similar in some respects, is quite different in that it seeks to establish new communities rather than re-establish old ones.

The "personal discovery" sites face another constraint which will always limit their growth potential - we have already seen the passing (in the social networking sense of the word, anyway) of the last pre-email, pre-mobile phone generation. From now on, the new cohorts of users quite simply will never face the situation of having lost touch with someone against their own will. They grow up permanently online and on-call. In a sense, this is quite sad - if you've never lost, you've also never found. But that's just lyrics, I guess.

The polar bears can rest easy, for now...

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

The Stockman gas condensate project goes on hold, as investors lose c.$400 million each

Remember the"war of words" over mineral rights in the Arctic a few years ago? The story went like this: while the commercially viable reserves elsewhere are peaking or declining, the Arctic harbours untold riches of oil and gas, which are now becoming accessible due to global warming. Therefore, all the countries with an Arctic coastline (the United States, Canada, Denmark, Norway and Russia) felt obliged to grab as many square miles as possible. Now it looks like all the excitement was a bit premature.

For many years, one of the biggest potential E&P projects in the Arctic was the Stockman deep-sea gas condensate field in the Russian sector of the Barents Sea. The investment thesis was based on bringing the gas to shore, liquefying it there in an LNG plant, and transporting the LNG by tanker to the US (at the time, Henry Hub prices were quite tasty). Initially, Gazprom, which holds the exploration licence, planned to go it alone. Then, after the scale of the capex requirement and the technical complexity had become clear, it invited western oil companies to come in as partners. Eventually, Norway's Statoil and France's Total joined the consortium, with Gazprom keeping 51% of the interest.

For years, things have been suspiciously quiet, and now we know why. The consortium has been officially disbanded, the three partners having lost over $1.2 billion between them on feasibility studies and such like. The Statoil representative has insisted that the project remains technically feasible, but "the commercials needed to be better defined". What it looks like is that the cost estimates have spiraled beyond the parties' willingness to invest, at least for now. I also suspect that the decline in US gas prices, due in no small part to the increased production of shale gas, has also played a major role.

With this flagship project hitting the buffers, expect the "race for the Arctic" to slow down considerably. Incidentally, this should give the US an opportunity to re-enter the fray in a much stronger position. Unlike the other four "Arctic nations", the US is not at present a party to the Law of the Sea Convention. Should the US sign and ratify it before the other players take a fresh look at the opportunities in the Arctic, it will become a major force to be reckoned with - in this area, it is not one at present.

BitTorrent and the music business: an uneasy truce emerging?

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

File sharing is here to stay - music companies need to learn to live with it

I've come across this interview with one of senior BitTorrent employees, Matt Mason, who has written extensively on the issues around music piracy. BitTorrent, in case one doesn't know, is an incredibly powerful tool for P2P sharing of high volumes of data. As such, it has succeeded where Napster failed before: it has made music (and video) piracy more than technically possible - it has made it virtually irresistible. Does this mean that the music recording industry is finally doomed?

In my view, what is doomed is the music divisions of the huge US studios, which may well follow  EMI into the sunset, unless they radically change their game. For we are now seeing a likely end of paid content. This is a much bigger challenge than that posed by "first generation" P2P file-sharing around ten years ago. Back then, the CD was king, and the king was way over-priced! As a reaction to that, young people started ripping their CDs into mp3 files and swapping them over the likes of Napster. The trade-off was poorer sound quality and a lack of a  physical medium (although of course you could then burn your pirated mp3 tracks onto a CD...). What that meant was that, having found the music that they liked via free file-sharing, people were then willing to go and buy it in physical form, through an "official" retailer (and it helped that, faced with the threat of piracy, CD prices declined to more reasonable levels). The music studios' revenue model therefore had to adjust, rather than crumbled completely.

What happened then, though, was that the studios, and of course Apple, saw this as a opportunity to get rid of the physical medium altogether. Of course this reduced the distribution costs and sped up the buying process. And of course Apple made a lot of money. But once a professionally-produced and professional-looking physical medium is no longer the default, there is no longer any difference in the user experience whether you go through the "official" or "pirate" channel to get your music tracks (or the latest episode of Hardcore Pawn). In fact, ownership is no longer viewed as very important, since with high-speed broadband, 3G (and soon 4G), wi-max and other "always on" technologies, consumers can simply stream whatever piece of content they fancy this very second. In this sense, perhaps the bigger challenge to the studios is not BitTorrent but services such as LastFM and YouTube (although they do pay royalties to the rights owners).

But back to BitTorrent. What Matt Mason is saying is that the studios can monetise their content rights by using P2P portals as audience aggregators for the purposes of advertising selling, product placement etc. Facebook has shown how this should be done. Of course, Facebook's challenge now is that it is no longer seen as "cool enough" by a lot of people. BitTorrent and similar services should be immune to such a "falling out of fashion", as their focus is external (music content) rather than internal (interaction with other "cool people"). Don't treat the file-sharers as pirates, treat them as consumers (and milk them accordingly!). If that's how it pans out, the recorded music business model will come to resemble that of FTA television, in being predominantly advertising-funded. Isn't it curious how the future of one medium may well lie in the past of another. But before that happens, a lot of companies that once seemed "too big to fail", will have failed. And hopefully their place will be taken by a large number of independent labels, or even artists marketing their music direct. But that's the topic for a separate article.

Back to the gold standard? You are having a laugh...

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

The US can't afford to go back to the gold standard - it needs fiat money to service its huge debts

It's a truth universally acknowledged that the international financial system is broken is some fundamental ways. Many blame the huge increase in money supply (and, linked to that, of cheap credit) since the early 1990s for the over-heating of the global economy that culminated in the 2008 crisis. Since then, while the full economic collapse feared at the time has been averted (at least for now), many of the leading economies are suffocating under the weight of debt, accumulated both in the run-up to 2008 and in the "quantitative easing" afterwards. In many senses, the cure has been worse than the disease.

Against this background, the Republican Party is considering returning to some kind of a gold standard system for the greenback. While the stated commitment to a saner monetary policy is laudable, is the suggestion itself a credible one?

Let us remind ourselves what happened in the 41 years since the US abandoned the gold link and introduced a fiat currency. In 1970, the last year under the gold standard, US federal debt was 28% of GDP, whereas for 2012 it is estimated at 74%. With the individual states included, the public debt was 102% of GDP in 2011. Of the industrialised countries, only Japan and PIIGs had more. Also, given the size of the US economy, we are talking some truly astronomical numbers. In a normal economy, this mountain of debt would result in pretty high interest rates, leading to a full-blown depression (as it has done e.g. in Greece and Ireland), and possibly riots in the streets. In reality, though, 3-month Treasury Bill yields have been hovering just above zero since 2008, and are projected to remain there for the next couple of years.

The reason for this is the liberal printing of fiat money by the Federal Reserve System. Unlike Germany in the early 1920s, hyperinflation has not been the result because the US dollar is the world's main reserve currency, so foreign investors have been happy to soak up trillions of freshly printed dollars to buy Treasury Bills, still considered to be a "safe heaven". Of course, this leads to escalating debt, which needs to be serviced, etc. Eventually the bubble will burst - but not quite yet. Returning to the gold standard would prick the bubble immediately and bring the whole sorry edifice crumbling down. Even the most ideologically-driven Republicans are not "brave" enough for that.

The Murdoch girl and media ethics

This post was written in 2012–2013 and reflects thinking at the time. For current views and topical discussions, please see recent articles.

The lady doth protest too much - Elisabeth Murdoch is an unlikely example in media ethics

So, Elisabeth Murdoch has been taking pot shots at her father and brother. Apparently, "television is a force for storytelling rather than a route to political power" (implying that nasty old Rupert has been subverting British democracy). Also, she is unhappy that News Corp is being run by people fawning over the boss rather than being governed by a "rigorous set of values". Of course, The Guardian just loved it - and wouldn't any fair-minded individual?

Actually, no. First off, there is a stench of hypocrisy emanating from Elisabeth's sermonising. Where would she, and her "independent" media production company Shine, be without News Corp's backing and her family's industry connections? Didn't she make untold millions by selling Shine to News Corp?

Secondly, television is (or can be) many things, including "storytelling" (whatever she means by that). But it doesn't have to be very narrative in nature - minimum-commentary rolling news coverage also has its place, and so does mindless entertainment (such as much of Shine's output). More to the point, "storytelling" is not by definition a good thing - we've seen many examples of mendacious stories being rammed down the public's throat by biased or government-bullied media organisations ("Iraq's weapons of mass destruction", anyone?). You can't accuse e.g. Fox News (one of Daddy's companies) or CNN of lacking a clear narrative. (This becomes especially apparent in any foreign conflict in which the US is involved, from Yugoslavia in the 1990s to Libya and Syria today). So a propensity for "storytelling" only becomes a virtue in a media organisation if there are equally vocal organisations espousing alternative stories. All too often, though, that's not the case.

Thirdly, you can't deny that television is also a route to political power - always has been, always will be. Some media owners are more subtle about it, others less. And that includes dear old BBC and The Guardian itself. (If anything, my impression is that in the UK News Corp has been less politically activist than in some other countries). So Elisabeth is offering a false dichotomy, which isn't helpful.

The Guardian article mentions the idea that the motivation behind Elisabeth's speech is trying to get the Beeb to commission more programmes from Shine. That may well be part of it, but perhaps sibling rivalry has played at least as big a role?