As Reuters has reported within the last hour:
Japanese technology shares fell on Tuesday as a global market rout sparked by the emergence of a low-cost Chinese artificial intelligence model entered day two.
The market value of the much-hyped US AI tech stocks has tumbled by more than $1 trillion in a day. Apparently, shares in AI chip maker Nvidia fell in value by 17% on Monday. That reduced its value by $593 billion, which was a record one-day loss for any company.
What's the reason for all this? It is that a relatively unknown Chinese AI engine called Deepseek, which basically undertakes the same tasks as ChatGPT, has released its tech on an open-source basis.
That, however, is not the big threat because, by itself, that would not threaten the value of the vast investments the US is making in AI. What does that is the fact that Deepseek apparently delivers results that are as often as good as those of ChatGPT and its like, and does so using vastly less computing power and so vastly less energy. That is a massive game changer in the AI market, if true, overcoming one of the biggest issues within it and the objection that many have to its use.
Even more so, though, Deepseek does something else if the claims made about it - which appear to be well-founded - are true. It challenges the foundations of modern US capitalism.
US capitalism has three core ideas at its heart. The first is that ownership is king. Protecting whatever is created from replication is the first tenet that it follows. Nothing is more important to it than. Concentrating the power of ownership in the hands of a relatively few people is all US capitalism is about now.
Second, as a result, secrecy prevails. The patent lawyer is the architect of business success in these enterprises - by securing the knowledge of the organisation from attack. Massive barriers to entry for competitors are built and vigorously defended, completely contrary to the ideas of free markets.
Third, the goal is to maintain this situation for as long as possible, allowing for the maximum extraction of profit from the consumer through extortionate rents. The consequence is that US capitalism is frequently bloated with high costs, which only appear reasonable because of the extortionate fees monopolistic rent-seeking entities, protected by vast legal infrastructures, can charge.
In summary, then, big US business is inherently anti-competitive by nature, is monopolistic by choice, and is, in most cases, very likely to be grossly inefficient as a consequence, with the costs being borne by the consumer.
What Deepseek has really done is challenge this model. It is low cost. It is efficient. It has been created in an intensely competitive environment, and to win competitive advantage through widespread adoption, it has opted for an open-source model, which is the complete antithesis of what a US corporation would do. This is profoundly disruptive to the US way of doing business, and because Deepseek apparently works (although it has been so inundated with requests for accounts I have been unable to get on as yet), it is very unlikely that its strategy will fail. US capitalism has, in that case, met its nemesis, and the fact is that when doing so, that nemesis is doing exactly what economists say should happen in free markets when everyone knows it does not at the top of the US capitalist tree, which long ago rejected that idea of free markets and opted for one of exploitation instead.
The fact that this has happened a week after Trump gets to office, supported by Big Tech, cannot be ignored. Maybe the Chinese state has been waiting for this moment. However, Deepseek is not a Chinese state company. It is privately owned, albeit it complies with Chinese state requirements. The suggestion has been made that it is not too good on some questionable moments in recent Chinese history.
The point is that whatever the cause for the timing, the disruption in US capital markets has arisen for very good reasons. If China can develop efficient tech at low prices and put it into widespread use through open-source diversification while maintaining a fee base for access to more advanced searches, then the whole myth on which the next generation of US wealth extraction was being built has been shattered.
I am not a fan of the Chinese state. It perpetrates far too many human rights abuses for that to be possible. But if this company breaks US monopoly power, I will not be complaining. The US should worry, though, as should Trump. Its whole edifice might tumble if its tech power is shattered right now because the myth of US growth and stock market wealth is built on that one sector.
This might be a seismic moment.
And for those who wonder what Deepseek can do, look at Stephanie Kelton's post on Substack yesterday. In 12 seconds, it built a model that replicated the Bank of Canada's estimate of the cost of US sanctions on that country, with full reasons given.
Thanks for reading this post.
You can share this post on social media of your choice by clicking these icons:
You can subscribe to this blog's daily email here.
And if you would like to support this blog you can, here:
I was astonished by Stephanie’s post yesterday, particularly by the clarity of the workings and the way it produced an answer. This looks to be quite a moment although perhaps not so good for my pension!
Yes, showing the “reasoning” process (the output of each iteration of the algorithm as I understand it) looks to be very useful to allow better verification of the results and possibly as a tool to help people learn.
I have seen examples where the output took considerably more than 12 seconds (on the order of several minutes) and required several prompts but the output has been deemed at least as good as competing models.
These are interesting times indeed and I wonder how much of a coincidence was the unveiling of Deepseek so soon after Trump took office.
I am getting my head around how to write AI ‘prompts’ to get what I want. It is not hard and delivers impressive results.
Richard, excellent skill – ask it to produce a model and run a simulation to explain Rachel Reeves’ motivations and thinking in the light of her economic moves, and the likely impact of those moves in the context of the Labour Party Manifesto 🙂
Salk, inventor of the first Polio Vaccine poo-pooed the idea of patenting it
Davey, inventor of the Miners Safety lamp – George Stevenson also had a version didnt patent it because it would save lived
More importantly perhaps Malcolm McLean who put together the modern Container system – and bear in mind that he didnt design it himself, he had to involve countless engineers and designers not only to design the containers but their fixings, stowage systems, ships, lorries and railway wagons to carry them etc etc. didnt patent the system but instead gave it to the world for nothing.
Without his idea there would be no world trade in the way we have it today
Can we add Volvo cars invention and development of the 3 point seat belt to your list. Has saved countless lives across the world.
“Without his idea there would be no world trade in the way we have it today ”
This being desirable? – vast amounts of trade? The UK de-industrialising and exporting emissions to China? (ditto the USA and others).
Or: the massive proliferation of drugs hidden in containers (Rotterdam, Antwerp, Hamburg).
Just saying.
George Stephenson, not George Stevenson, John.
Stevensons built lighthouses apart from Robert Louis who wrote damn good books. Useful family. And I’m not related a far as I know.
There are two additional points I would make, after acknowledging that the panic we have seen in markets is not a proof that we are dealing with established facts; markets panicking is rather a market only of the special ability of markets to panic.
Assuming this is what it a appears, there are two observations i would wish to make; first (as I commented in a previous thread), this is an excellent illustration that real power, real economic change does not come from politics, but from innovation; especially applied innovation, most readily in technology (here, Korotayev for example, provides useful insight). Second, if it is true it makes China a real imminent threat to America. DeepSeek (wiping £1Trn off Big Tech share values) would follow TikTok, and BYD (already beginning to out-perform Musk-Teslar in electric cars). Meanwhile Trump is tying the US to the mast of old technology (drill, and drive 4-stroke cycle cars); and that is a sharply distinct economic profile. Which one is the future?
Good questions, John
I also ask this question; DeepSeek is Open Source (now downloaded round the world in mind-boggling numbers), and from China. Who is DeepSeek? Who owns it? Who developd it? Who funded it?
Tbis is all in the FT today – but I do not have time to grab the data now, or the rest of the day – which is going to be very busy
Good points. 2019 Bx: me meeting with VW (nice view in their offices – overlooking the Berlaymont). Tesla invited some VW engineers to USA – they came back frightened – Tesla doing with 50 engineers what requires 500 in VW (this is 1st hand btw). Current EU situation: I can buy a new EV in Ukaraine (Chinese or EU) for circa Euro20k – same here in EU – Euro50k. This is not just due to VAT (no VAT on EVs in Ukraine) or low/no import taxes. The Euro OEMs (Original equipment manufacrturers = the car mfus) want to keep their show on the road – EVs take bugger all in servicing (what’s to go wrong?) so you have to make the money somewhere. What do the Euro OEMs do? lobbying to keep ICEs (internal combustion engines) beyond 2035. How does that work when at least one Chinese OEM has an EV that will do 1200kms between charges? & it can cost peanuts to charge it. Politicians – one side of the Atlantic – dumb (Trump), otherside: bunnies in the headlights of a truck. This starts to look like wipe out – millions of workers (in Germany) out of work & the politicos don’t have a clue. Rinse and repeat in other areas.
Scarey….
This has to some extent pricked the bubble in some tech share prices, which have increased beyond all reasonable metrics. If this slide continues for another few days, might this be a dotcom moment?
If it continues…., yes
“IF”.
I do think this is an important moment in the technology industry where China, once seen as delivering second rate, poor quality, copies (in a way that Japan was perceived in the 1960s) can now compete with the US (as Japan did in the 80s and beyond).
The broader US stock market has yet to really move much. S+P (futures) fell from 6,150 to 6050 (now) – which is still much high than 2 weeks ago (5,810) and election day (also 58,10). It has been far more interested in interest rate expectations.
Whilst “Cheap AI” is bad for NVIDIA it is good for the rest of us (well, at least better than expensive AI!). Trump is right (never something I thought I would ever say) – US players will have to sharpen their game.
USA / the West have a big problem as they have adopted neoliberal economics of financalisation that has decimated their industrial base in search of profits by off-shoring industry. It will take decades to rebuild real GDP with real industry but a GND Green New Deal should be the priority including a roll out of local public transport eg. electric trams and buses. The UK has some good VLR Very Light Rail developments …VLR Coventry , Trampower, SET inwheel motors and steerable axles but as always governernment funding is lacking for this long horizon payback. Private investment = short term payback Government investment = long term payback
In the dot.com boom, the City was throwing multi-millions at twenty-year-olds, who had never run more than the proverbial paper round, in technology that nobody in Venture Capitalism or the City knew anything about, but still threw the money around in case they were left out; and the clueless twenty-year-olds couldn’t make work. Most of them failed. Whatever you say about BigTech today; they are nothing like the dot.com non-business opportunity in 1994-5; a market that couldn’t distinguish between absolute rubbish, and Google.
Here is the irony. Google began as Open source, and idealistic. It was the dot.com bust that ended all that, and brought in Big Money, and everything that followed.
Releasing Deepseek a week into Trump and his tech buddies’ triumph can’t be a coincidence. It’s too funny.
I have great admiration for those who eschew wealth to bring something useful to the world. They’re a different breed.
Reading Stephanie Kelton’s post, however, I’m reeling from the shock at what this technology can do and the speed with which it can do it.
I can’t be the only one who feels unsettled and worried at the change this will bring to people’s lives – if I could work out exactly what those effects are but it’s beyond me!
Ask it…
LOL!
I’m so ignorant of the whole thing I never thought to.
My brief experiments yesterday left me with the distinct impression that DeepSeek is significantly more impressive than the publicly available American offerings. If it’s true that it is also less power-hungry, I think it will be truly disruptive, not just of the US tech sector but in the wider uses of what it can offer.
Hello Deepseek, please let me have your revised staffing projections for the Bank of Canada.
We have been here before.
1987. European governments decided that they needed a EU-wide (open) mobile phone standard. It was written (mostly) by “Groupe Special Mobile” (GSM). It mostly defined intefaces and functional blocks within the system. It was an open standard. Much of the work was done by… state-owned telephone companies (= strange but true).
Develoment was fast & the 1st GSM network went live in Helsinki (natch) December 1991. it conquered the world & for +/- 14 years the Euros ruled the roost in mobile telephony/mobile data comms. The USA then got its act together (Apple) ditto the Chinese (Huwawei). GSM was successful becuase it was an open standard & the Euros lost control due to failure to move fast enough.
Something tells me the Chinese won’t make the same mistake.
PS: the DeepSeek development has implications for Taiwan since it opens up possibilities for mobile/drone-based A.I.
If people want to use these models to attempt to develop ‘smart’ drones, any sort of an ‘AI’ engine could be used. Deepseek, ChatGPT, etc. All open source, all can be developed by anybody with the correct know how.
I suspect that the military (more precisely, the large weapons manufacturers) are going to be using more specialised models, perhaps developed from one of the open source models. And they’ll be inextricably linked to their expensive drone hardware, of course. The arrival of Deepseek won’t make any difference to this, one way or the other. Where there’s a will, there’s a way, as the saying goes.
Here’s an interesting video in which a former Microsoft engineer explains the difference between Deepseek and the other well-known models:
https://www.youtube.com/watch?app=desktop&v=r3TpcHebtxM&t=8s
It’s a clever enough idea, but we’ll see ChatGPT/Gemini/Apple/Grok etc etc etc copying this approach themselves within weeks.
Thanks
While the markets have given huge valuations to the likes of OpenAI and Nvidia, some like myself see those valuations as flawed.
I’m working with AI daily, and one thing that’s always been clear with LLMs is that there is little lock-in. The applications of AI that are verified, proven, and which require niche knowledge to create, WILL have substantial value in various cases. But LLMs that struggle to earn enough to cover their costs which can be swapped out trivially have always had this risk.
Similarly people have seen the escalating power usage and feared for the impact while missing that performance for a given power usage had kept increasing rapidly month by month, so once a suitable performance could be achieved then the cost and energy usage could be shrunk instead.
There have been applications which could switch to the Mini model or Llama or whatever already, Deepseek was just the same thing done again compared to O1 and not even the first to approach it given QWQ came out first.
It bodes well for local LLM prospects and various investments that need high quality results at an affordable level. OpenAI may well worry. Nvidia, too, especially given a number of those running Deepseek have been using CPUs. In any case hardware innovations including both dedicated LLM chips and even analog encodings for vastly better performance/power will further challenge Nvidia in due course.
An excellent summary of the implications and much to agree with.
I cannot however ignore the unease i feel though with these applications.
And it’s not just in AI, Richard. About four weeks ago there suddenly appeared in the skies of a certain part of China two versions of a stealth fighter/bomber. It had been believed that the US was years ahead of the Chinese with this kind of development, but it now appears not. For anyone who has an interest in military technology (something I’ve become interested in since the war in Ukraine began) – or just want evidence of my claims see this link: https://www.twz.com/air/chinas-mysterious-tailless-stealth-fighter-has-swiveling-wingtips
And it doesn’t stop there. I’ve long been anti Chinese cars – largely because I’m not a fan of the country, for the reasons you outline (but also because their claim to be a ‘socialist’ country is about as pathetic as me claiming to be King Kong). But as a bit of a car ‘bore’ I have to admit that reviews of many Chinese electric cars are very good – certainly as good as European equivalents, and almost always cheaper (rather as Japanese cars were in the 1970s to 1990s).
So, yes, I agree, the US monopolists have got a lot to worry about – hence Trumps animosity towards China.
Agreed
> it has opted for an open-source model, which is the complete antithesis of what a US corporation would do
To be fair, the usually evil Meta (formerly known as Facebook) have also released all their Llama models as open-weight models so far. Though in contrast to DeepSeek they have never released a model that was as capable as OpenAI’s latest at the time, they’ve always been about a year behind.
Also, this is being a little pedantic, but the term is open-weight rather than open-source. Plenty of others get this wrong, so I’m not blaming Richard. Open-weight means that anyone can download the weights and run them on local hardware, if they meet the hardware requirements. The term open-source doesn’t really quite apply to Language models but if it was fully open they would have released all the training data and fully documented their training methodology also. As far as I know they haven’t done this. Neither has Meta.
The hardware requirements to run the full-sized model are significant enough, requiring on the order of 10~16 top of the range NVIDIA GPUs, beyond the resources of most individuals but not that of many organisations. By opening the weights this way, it means that many people can now run their own SOTA AI model without having to share data with anyone, which opens up AI in a whole new way.
> The suggestion has been made that it is not too good on some questionable moments in recent Chinese history.
Apparently this is only if you access the model via their chat application. If you download the model and run it locally, by all accounts the censorship disappears, which strongly suggests that the responses of the model are not aligned with CCP policy but that the responses are post edited.
Thanks. Appreciated.
The bsic question we need to ask is; given Google was originally OpenSource, but is now a different, profit-driven model because in a dot.com crisis capital would not back it without a big payback; why is China offering OpenSource AI on this scale?
To disrupt markets, and secure market advantage.
I was thinking along the lines of the relationship between Chinese business and the Chinese State. I noticed in a thread (this or a recent one) a commenter dismissed the idea that China was really a Communist State. A big mistake. China has developed a unique Chinese form of Communism, reflecting the ageless unitary nature of the Chinese State and culture; devising a highly pragmatic form of State-backed Capitalism that discovered from bitter experience in the cultural Revolution that communist economic theory didn’t work in the real world, but, through Deng Xiao Ping in the 1970s-80s, realised that capitalism’s power is that it can work in almost any environment, and capitalism’s natural pragmatism fitted China like a glove (the Chines in Shanghai and Hong Kong had known that for well over a century). Deng Xiao Ping understood that capitalism is endlessly adaptable, capricious and protean – but manageable; just as long as the framework allows it to do business and generate profit (it will always chafe at the restrictions, but rarely willingly bites the hand that feeds it). And that capitalism throughout history has thrived best when nurtured and developed in close cooperation with the State. Capitalism did not begin with free market capitalism, and then prospered: the history is always of state support (of one kind or another) as the foundation of success.
Free markets and a small state more typically end in failure (a crash and depression), or prove high risk, and threatened by instability; or remain an ideological vision for the future, never fully realised. In no version of our world, does the US today, with the scale of its military-industrial commitments, represent “small state” capitalism; nor could it do so.
What’s happening reminded me of an old computing joke: If I had a penny for every dotcom share I owned.
I have many thoughts about this, and this is coming from a tech perspective, not economics. Open source as an “assault on capitalism” is by no means starting, it’s been around for a long time: Linux, VLC, (more recently services like mastodon and pixelfed), and many others have been challenging your 3 core principles mentioned for decades. But this may be the first time that the market has really stood to attention and reacted in such a viscerally negative way. Why?
I think because fundamentally there isn’t really a whole lot going on in large language models (and chatbots). It is very much a bubble. A massively overhyped and inflated market, with so much investment, to produce a tool with very limited use cases. Until now US markets have been able to coast on the magic and glamour of things like ChatGPT, but now a Chinese firm comes along and just pulls the rug from under them. They can no longer claim to justify the investment purely using hype and potential, so will now need to scramble to produce real profitability, which they are still struggling with. It’s easy to get millions of users when you offer a mediocre chatbot for free, once people and organisations start having to pay hefty subscriptions for access, things will look very different.
Investors have been strung along by US tech emperors throughout this whole LLM /chatbot boom, and now someone has pointed out they aren’t wearing any clothes.
I agree. We (me & others) are developping a Level 3 A.I. (machine learning) in the context of renewable power generation, storage and load on electricity networks. Because the generation is stochastic (day to day) albeit with forecasts for weather, the cleaver bit is to gradually learn from the forecasts. After that, I plan on developing the A.I. to auto-plan community energy systems wrt the renewable resource & based on that plus network structure and load to decide on optimal location of assets. Cuts out repetitive “modelling” and gets you quickly to where you want to be.
That is interesting.
I do that job every night at bedtime on my phone, to set up my storage battery, depending on local weather forecast for the next day, I have to decide how much off peak electric (7p/kWh) to buy between 12midnight & 5am depending on how much sun is predicted for the next day, and whether there is laundry to do, and whether the meal is to be cooked on the gas hob, roasted in the electric oven or reheated in the microwave.
This year I’ve noticed the local (Met Office) weather forecast is more “stochastic” than it used to be , predicting sunny or dull, and when, LESS reliably than it used to because weather IS more random than it used to be.
At present I have to use my “natural” intelligence. As it’s a Chinese system (NOT Tesla), maybe DeepSeek will give me a discount.
it has opted for an open-source model, which is the complete antithesis of what a US corporation would do.
But US corporations have been developing open source models. Google’s Tensorflow AI is open source, as is Musk’s grok.AI. And Point-E, Whisper, Jukebox, and CLIP are open source models published by OpenAI
But they are not where the money is going
The US stock market bubble is due to 7/8 public listed tech companies. The rest of the quoted companies are no where near the same valuation sums.
Realistically the bubble was always due to pop.
The hype about A1 and only the US can do it has been well and truly burst.
The arms race between US and China has truly begun in earnest and the US Tech Bro’s have plenty to worry about.
Words like seismic, game changer, monumental huge paradigm shift, historic, truly ground-breaking, dawn of a new era, revolutionary, unprecedented challenge to US/West based driven economic/society theory/models all capture this moment very well.
The ripple effects of AI are massive and will trickle down to all aspects of life on Earth. We must now carefully consider this brave new world because over a very short period we will witness changes that impact, education, health, jobs, research, society, world order, power, politics, religion, wealth, every strata of our being will be tested severely.
Overall, the launch of Deepsake has to be celebrated far and wide for all the reasons you have highlighted above. Not least because the impact of AI on the environment/climate will now be seen in the light of all the possibilities offered and and that have now opened up as a result.
Worth also mentioning these clips from Jack Ma
https://www.youtube.com/watch?v=rHt-5-RyrJk
https://www.youtube.com/watch?app=desktop&v=Dd5j7ccujew&t=61s
There’s pretty good evidence that China has already pretty much won the tech race – “China has a ‘stunning lead’ over the US in the research of 37 out of 44 critical and emerging technologies” – https://www.businessinsider.com/china-leads-us-critical-emerging-technologies-strategic-competition-research-report-2023-3
Thank you to Richard and the community. I’m delighted that Richard has posted and facilitated a discussion.
Further to comments from John S and Mike and Clive, may I add:
British and EU authorities have long been warned, to quote John S, that ” that real power, real economic change does not come from politics, but from innovation; especially applied innovation, most readily in technology”. They cho(o)se to ignore these warnings in favour of whatever Wall Street and City shills were / are selling. The UK is probably the worst.
Mike P: “Politicians – one side of the Atlantic – dumb (Trump), otherside: bunnies in the headlights of a truck. This starts to look like wipe out – millions of workers (in Germany) out of work & the politicos don’t have a clue. Rinse and repeat in other areas.” I have observed what led to Mike’s conclusion.
Clive: “Trump is right (never something I thought I would ever say) – US players will have to sharpen their game.” I must quibble a bit with Clive. It’s not just US players, but their government, too. It’s the same for Europe, including the UK.
From the G20 in March 2009, China and Russia began to second young government officials and bankers to City regulators and firms, learning what not to do. I was involved in their training. The sponsors were thinking and for the long-term.
A year or so later, a Belgian Green MEP, Philippe Lamberts, proposed curbs on financial services pay, so that Europe’s best and brightest would go into other, more productive sectors. He was laughed at and discouraged. China learnt from that, too. Russia, too, I would say.
Although paid to tout neoliberal BS, from time to time and when appropriate, I talk(ed) out of turn. There was / is no interest, not even asking why I hold these views.
I think this is as disruptive as the PC, and before that the hoover and washing machoine, and maybe car for most families.
I have now teried the model asking questions on obscure subjects of which i have some knowleldge. The LLM is impressive based on what it has given me.
Its interesting, especially if you are in the middle of it all to try and work out whats really ‘game changing’ and whats an old product rehashed.
After all E-Bay’s Exchange & Mart Mk2 while Amazon in its original form is just a mail order company who does it over the internet.
In the same vein it was never a done deal that cars were going to be petrol, they could have been electric – or steam in the early days. But of course as has been pointed out in this thread the real issue for the car industry isnt so much going from petrol and diesel to electric BUT the sudden crash in demand for spares and service as a result.
Interesting times!
(Anyone Fancy a Doble model E or F?)
This 2022 article explains the stupidity of both Biden and Trump in regard to playing fast and loose with tariffs. Indeed the large inflationary rise in house costs I’m sure played a large part in voters rejecting the Democrat Kamala Harris for presidential office:-
https://www.seattletimes.com/opinion/tariffs-on-canadian-lumber-are-driving-up-home-prices/
The article needs to be read by Brits knowing that the bulk of US homes are built using timber framing and roofs are fully lined out with particle board to receive bitumen shingles.
“ That is a massive game changer in the AI market, if true, overcoming one of the biggest issues within it and the objection that many have to its use.”
Objection to AI goes much further than energy cost, although it’s a significant issue, but I have a few other remarks in mind
Mainstream economics is an ideological weapon legitimizing tyranny.
Of the many policy measures deployed in plutocracy’s favor has been the criminalization of a basic pillar of innovation, which is the right to reverse engineer and then tinker with existing tech technologies. If you’ve ever noticed the trend of putting software in just about everything, that’s key. Whales grow by acquisition rather than innovation, yet the geniuses of the world have all but killed off the capacity to innovate in the first place.
The cumulative dysfunction in science and expertise along with demographic damage from 2008 combined with financial stress on regional colleges, spell an extremely bleak future for the US technologically. I suppose all of this is supposed to be papered over with H1B Visas and some magical artificial intelligence that’s going to replace human research.
The UK, is also suffering many of the forces. It has been a shared ailment starting with Reagan and Thatcher. I think people should appreciate the scope of this assault on the foundations of civilization.. It’s global scale insanity.
The DeepSeek watershed moment Why an open-source AI model from China is reshaping the global tech scene
The story of DeepSeek-R1 isn’t just about a single open-source model. It’s a signal that the AI landscape remains wide open. Constraint-driven innovation matters.
Whatever happens next, one thing is certain We are nowhere near the end of the road for large-scale AI breakthrough
The market and MPs councilors and Metro Mayors such Steve Rotherham Mersey Region Metro Mayor, barely understand underlying tech, and that goes even more so for AI/ML systems that politicians and the media can only grasp through second or third-hand information. The biggest amount of disinformation today is literally about AI as well as the literal output of LLMs that dominate AI conversations, which have absolutely no guarantee to be true and are often plainly false and should not be believed.
The biggest threat is not from China but from the private ownership of LLMs that will dominate enterprise IT and social media in a matter of months not years.
Some years I wrote a somewhat cynical & downright satirical piece about ‘cutting without bleeding’ – a 10 step guide to improving efficiency in govt services. Even Eric Pickles liked it…
Anyway – I thought step 10 seemed appropriate in this instance…
Above all, whatever you do, do not let anyone including even your most trusted lieutenants suggest that your assumptions should be examined. You are a senior leader and therefore all that you believe and say must be correct, if not enlightened. 500 years ago Machiavelli may indeed have suggested that true leaders need people around to tell them the truth. But he was clearly wrong as evidenced by the fact that he died a long time ago. Be certain, be sure and be confident: the strategy that you are implementing is unavoidable. (You may wish to write this last statement out and place it on your bathroom mirror.)
You can read the whole blog post here:
http://jonharveyassociates.blogspot.com/2010/06/cutting-without-bleeding.html
I haven’t read your post – but I take the tongue in cheek tone
May I suggest that the majority of these points boil down to the fact that, oh dear, China actually understands money…
Wow, wow and wow again. 12 seconds. When I studied economics in the 1970s this would have been days, if not weeks or months!