Techstorm AI Edition - January 2025
Part 2 of 2, going down the AI rabbit hole!
OMG, what a crazy year it’s been in AI. About 12 months ago, when I was fed up with all the AI hype, I sent out a somewhat ironic (cynical?) Techstorm Forecast on AI, including a Top 10 list of what was buzzing. I thought of this as a one-time thing, but recently, people have asked for a new AI Top 10 list. So here it is, yet another end-of-the-year semi-serious compilation of what’s in vogue in AI.
Generative AI is the tech roller-coaster ride of the decade. It’s not easy for anyone to keep up with what’s happening in AI, which LLM is currently the most potent one, the latest crazy valuation of OpenAI, whether their business model is sustainable, and how we are going to be able to provide energy to these power-hungry AI data centers.
This is my humble attempt to bring some sense, clarity, and perspective to the ever-inflating world of (generative) AI.
Although there’s a fair bit of irony in the AI Top 10 list below, don’t interpret this as a techno-pessimistic rant! What we’re seeing is short-term Generative AI hype that is, unfortunately, clouding a huge long-term AI opportunity!
Download the forecast, install as lock screen image on your phone, and use it to sound really smart in your next C-suite meeting. ✌️
#1 on the list is Generative AI, on the one hand, full throttle with tons of VC money pouring into every company that’s only somewhat related to AI. On the other hand, we’re seeing quite a few technical challenges when developing and running these systems. Not only is the area facing setbacks when training new models, but the improvements are actually not that impressive. The Gen AI industry is definitely in strong need of some re-invention when it comes to both training and running these LLMs. Maybe a below-the-radar Chinese startup has found the Holy Grail of Generative AI training? Looks quite promising, if you’re willing to take the risk of having a Chinese LLM provide input to your decision-making/coding/writing/cat image generation.
Then on #2 we have Nvidia, the actual winner of the AI race so far. The stock is up 171% in 2024, and even the company janitor is now a millionaire. It’s fair to say that Nvidia is an awe-inspiring company that has worked for decades to get to where it is today. It’s also fair to say that strong FOMO vibes from VCs, startups, and Big Tech drive some of this recent growth. Time will tell if it turns out to be yet another Cisco moment.
On #3, we have the AI Bubble, oh so present and inflated. Just don’t confuse it with the long-term prospects of AI, which is real but will take time. The bubble is instead very much related to Generative AI, the hype around inflated expectations, productivity gains, and excessive amounts of investor money being poured in. No matter how hard I try, I can’t see justification for many of these investments.
Sam Altman is dropping a few spots on the list to #4. Partly because other AI people (men) and LLMs are competing for the spotlight and partly because of the revolving doors on the OpenAI executive floor. ”Of 13 people who helped found OpenAI in late 2015 with a mission to create artificial general intelligence, only three remain.” Of those three, one is on a sabbatical.
New on #5, an honorable mention to Demis Hassabis and the Google DeepMind team for sharing the Nobel Prize in Chemistry in December. Now’s the right time to show the world the real and valuable applications of AI, not just improved ad targeting, cold-email-outreach generators, and automated (generic) LinkedIn posts.
One new entrant in the Top 10 is Gen AI Agents, the next big thing for 2025 according to your favorite AI influencer. Will autonomous systems capable of sensing, learning, and acting upon outside input be the much-needed development that will take Gen AI mainstream (and profitable)?
Also new on the list is the urgent question of how to power these electricity-hungry systems. Estimations say a ChatGPT prompt needs 10x more energy than a Google search. Does the quality of the reply justify this, especially when considering that you quite often need to Google the GPT response to search for hallucinations?
With Google and Microsoft acknowledging that they’ve increased their carbon emissions by 48% and 30% in the last five years, it’s no surprise that they’re investing heavily in nuclear power.
So where are the consultants, you ask? Side-by-side with investment bankers and LinkedIn influencers, they’re working hard to position themselves everywhere on this list. And they’re doing this quite successfully, I must say. There’s no CEO around who can ignore the need to be digital-first and AI-enabled these days, at least if they want to please investors.
Beware that these experts sometimes change their opinions instantly to serve their sales desks and IPO teams. Read the Goldman Sachs critical Gen AI analysis from June and their updated positive Gen AI analysis from September if you need proof. That said, it’s perfectly OK to change one’s mind, but maybe not this quickly and this much.
A few things on my radar for 2025:
The generative AI bubble 🤖
The Magnificent Seven 💾
The economy, stupid! 💸
Quantum influencers 🎲
The German election 🇩🇪
The space race 🚀
Donald Trump 🍊
Populism 🤡
Europe 🇪🇺
China 🇨🇳
India 🇮🇳
USA 🇺🇸
That’s all for now, comments and social shares always appreciated.
Embrace curiosity with a bit of skepticism!
... Nicklas
One more thing …
It’s always refreshing to see actors play unexpected roles, like Billy Chrystal in “Before”. In this thriller series on Apple TV+, he plays a child psychiatrist fighting his demons while trying to help a troubled young boy. Clever script, excellent acting, and an honorable mention to Jacobi Jupe for playing Noah very convincingly.