Domestic business and investing titan, Dermot Desmond, upset the orthodoxy this week. Ireland’s 500-year plan to build the Metrolink might be cut short, even ended. Desmond suggested the €12 billion urban rail project due to start in 2028 could be a white elephant project superseded by AI and autonomous-driving vehicles. Any bets on the kilometres per annum build speed on this 18 kilometre ‘monster’? Actually, don’t bother. Reflect on China’s average motor expressway construction build of circa 8,000 kilometres per year. Then think about the UK adding barely 65 miles of motorway over the past ….decade. Given the Irish public service obsession with tracking the UK National Health Service or UK Housing/Planning as benchmarks, one shudders to think what our ‘ambition’ could deliver in over-spend and century-shifting deadlines. On a more positive note, AI could be one of the tools which could dig us out of our transport infrastructure black hole.  A bit early to call that one you might say, but I’m beginning to think another crucial economic sector which gets its fair share of criticism is enjoying the halo AI effect. Don’t bank on it but the banking sector is suddenly looking interesting…. The ”animal spirits” of Wall Street and record financial market highs always help the banking sector. Indeed Wall Street’s banks have just finished reporting quarterly results where trading revenues clocked a whopping $34 billion in Q2, up 17% on the previous year. Yes, the phenomenal gains in AI-focused stocks like Nvidia and Microsoft inflate bank trading revenues and drive increased investment activity but there’s more going on. You might have read about meme-stocks and unheard of companies in the US smaller cap markets (Russell 3000) tripling their share prices since April; 33 companies at the last count and only 5 actually making profits. But, banks as meme-stocks? Really? Well check out the Financial Times headline this week: “European banks get their meme-stock moment” Not even US banks, but European ones tracking an economic bloc getting its tummy tickled on tariffs by the Fiddler on The Roof of the White House. Can’t wait for the South Park treatment on that one, but back to the FT and European banks. When French banks like Societe Generale see their share prices increase by more than 100% year-to-date then my “spidey sense” tells me this is not about mundane cyclical banking drivers like trading revenues, interest rates or the shape of the bond yield curve. The aggregate European bank sector is up a whopping 40% in 2025 and there could be an (infra)structural driver of this story. Think back to our earlier sniping about Ireland’s struggles on transport infrastructure. Banks have struggled with unwieldy data and service infrastructures which have been a nightmare to upgrade to modern customer expectations. As we have written many times on these pages, the banks sit on some of the richest consumer data on the planet. Critical information on individual and institutional funding, spending and income patterns are in the possession of the banks. What if that data could be mobilised in a far more efficient way using AI and its agentic tools? Like Dermot Desmond’s thinking, could AI allow banks to skip an infrastructure bottleneck? It is early days but let’s take a look at a company you’ve probably never heard about before. Palantir Technologies might be named after a Tolkien crystal ball but it looks like its future might be right now, thanks to AI. The Denver-based company has been around since 2003 and specializes in software to analyze or “mine” data. Its early customers were government departments seeking assistance with unwieldy datasets and looking for actionable information. In particular, it gained traction with security/police departments searching for surveillance and predictive intelligence solutions. Sound familiar, or creepy? Park that thought and think banking. Then consider Palantir only just hit quarterly revenue run rates of $1 billion in its most recent results. However, that was enough to make it one of the 20 most valuable companies in America. Stock market investors think it’s worth $440 billion which is bigger than the mighty healthcare player, Johnson & Johnson (J&J) and its 138,000 employees. Yes, if you were wondering if the valuation of Palantir was looking a bit punchy, you’d be correct. Annualized revenues of just over $4 billion (vs J&J’s $85 billion) means the Palantir valuation multiple is currently 110x current revenues. The excitement and valuation is driven by two recurring messages whenever Palantir is mentioned:  

1. AI is accelerating the monetization of data infrastructure
2. AI is reshaping enterprise software and Palantir is uniquely positioned

  Palantir is expanding beyond government into commercial sectors like healthcare, finance and energy. The first thing that should strike readers about government and these three specific sectors is that they have enormous customer/user bases. This is the banking sector clue, and possibly its infrastructure B-AI-L out. AI will very likely remove the need for “transition” projects to upgrade data infrastructure and provide banking organizations with valuable action prompts which might even be carried out by AI-agents/bots. That’s a business model 'Hail Mary' for the bank sector and Wall Street’s banking analysts are doing something unusual too. Typically, bank analysts stick close together and move their recommendations in tandem with their competitor analysts at the other investment banks. Remember, “nobody gets fired if we are all wrong” is an established career strategy for the average analyst. This also means that share price targets set by analysts move in relatively small increments so as not to spook the herd or attract excessive attention to their analysis or models (usually flawed as with all human forecasting exercises). So, I was checking a few market analytics dashboards today and spotted the following: KeyBanc target price moved UP from $60 to $100 RBC Capital  target price moved UP from $63 to $97 Raymond James target price moved UP from $79 to $95 Believe me, 25%-65% banking share price target upgrades are not the done thing on Wall Street when TACO Trumpolini is threatening the Chairman of the Federal Reserve Bank on interest rate policy.  So, this is yet another sector to add to your list where the two letter response to any share price move query can be “AI”. However, at a structural level, you don’t need a Tolkien crystal ball to know that technology can transform the commercial prospects of a country or sector saddled with a perceived long-term ‘challenge’. I’m old enough to remember the gloomsters telling us Ireland was destined to perpetual under-development because we had no energy resources and could never compete in manufacturing/building things. Who knew? Maybe, the leaders who finally gave up on Ford in 1984 after welcoming and watching Apple begin manufacturing in Cork in 1980…..