Member Info for Dorfan

MV - ASI can’t do simple yes/no. I tried that and then got a 1,000 word reply, yawn!

ASI - you are now confirming that your use of AI is the definitive, correct and perfect, which is what I alluded to earlier. To me and others it comes over as arrogant and condescending and lacks the respect you are craving for.

ASI

“ Cutting out the cheap quips and throwaway shots would go a long way. In my world, respect is earned”

And people coming over as arrogant don’t get respect either, have a think on that one.

ASI - Question: can AI make mistakes and can it be trusted? Perhaps the best bit of this response is the ‘Bottom Line’ at the end. Perhaps you need to add to all your replies that the response potentially has mistakes and may not be totally trusted.

Yes — AI can definitely make mistakes, and the question of whether it can be trusted depends on how it’s used, what it’s used for, and who is overseeing it. Let me break this down:

Why AI can make mistakes

Training data limits

AI learns from the data it’s trained on. If that data is biased, incomplete, or outdated, the AI may give wrong or skewed outputs.

Misunderstanding context

AI doesn’t “understand” in the human sense. It predicts patterns based on probabilities, which means it can sound confident while being factually wrong.

Ambiguity in input

If the question or instruction is vague, AI may guess — sometimes incorrectly.

Technical errors

Bugs, misconfigured models, or insufficient compute can also produce errors.

When AI can be trusted

AI can be relatively trusted when:

It’s used in narrow, well-defined tasks (e.g. spelling correction, detecting anomalies in medical scans, optimising logistics).

It’s validated by humans — outputs are checked by experts before action is taken.

It’s transparent and explainable — users can see why the AI reached a decision.

It’s continuously monitored and updated with fresh data.

When AI should not be blindly trusted

High-stakes decisions without human oversight (e.g. medical diagnosis, criminal sentencing, financial approvals).

Areas where human values, empathy, or ethics are central (e.g. social work, counselling).

If the AI system is a “black box” and neither developers nor users can explain how it reached an output.

✅ Bottom line:

AI is powerful, but not infallible. It’s a tool — useful when paired with human judgment, dangerous if relied on blindly.

ASI

“ When I use AI, it doesn’t just regurgitate facts—it verifies them, expands on my input, and refines the language to make the message clearer and more digestible for the reader.“

And that’s my point, when you use AI your message is that we all have to accept it as the definitive answer and there is no need to reply back. Yet when I use AI you then challenge it to find holes in it to be factually incorrect to meet your own narrative.

With regards to talking past each other, your style comes over as talking ‘at people’ and not being respectful.

ASI - as per usual, hiding behind AI and thinking it defuses the situation and makes you feel superior again. The odd occasion when you interject with a personal comment and then get a negative reply back which you then find difficult to put you hand up and say ‘yes I got that wrong and you are right ‘ results in resorting to hiding behind AI to try and get you out of the cr ap.

ASI - so you are saying my AI is wrong and only anything you say by AI or otherwise is factually correct and we have to accept it and not contest it.

ASI - that is all complete conjecture, show me proof of anything that supports those figures and perception it is going to be reality. Have you ran it through your AI friend to see a reply?

ASI - let me just pull out the section of my AI which focused on ‘training’ 1 million people.

“ The training of “one million people” is promising, but “trained” doesn’t always equal “employed in well-paid AI jobs.” Some of that training may be introductory or upskilling, not always full AI engineering. And there is a lag between training and meaningful employment”

And the rest of the AI reply….

Where profits will go / who gains

Microsoft is a multinational headquartered in the US. A lot of revenue generated via usage, cloud services, software, IP, etc., will flow back to the company’s global / US base.

UK will benefit via taxes (corporation tax, possibly local tax or investment incentives), jobs, infrastructure, skill development. But the profit distribution (i.e. how much remains in UK vs flows abroad) will depend on things like tax policy, ownership of assets, licensing/royalties, and where the digital services are sold.

There is always a risk that much of the profit (especially from downstream usage or global deployment) ends up accruing to Microsoft in the US, unless there are specific agreements, taxation, or regulation ensuring value capture locally.

Skill / labour shortages

Even though Microsoft plans to train many people, there is existing evidence of digital skills shortages, especially in AI‐adjacent fields. That means some roles may have to be filled by imported expertise, or people trained abroad, or firms hiring internationally.

The training of “one million people” is promising, but “trained” doesn’t always equal “employed in well-paid AI jobs.” Some of that training may be introductory or upskilling, not always full AI engineering. And there is a lag between training and meaningful employment.

Overall verdict: Is the statement “fair”?

Yes — overall, it is fair to say Microsoft is investing billions in the UK for AI. But it is also fair to say that the sceptical concerns are valid:

In the short term, for people needing work now, the benefits may be limited, unless they already have or can quickly acquire the relevant skills.

The number of mass employment opportunities is unlikely to match the numbers of people currently unemployed or seeking work across many sectors (unless there are big complementary policies).

On profit retention, much will likely accrue outside the UK, unless strong policy measures are in place.

What would need to happen for more fair distribution/sooner benefit

To make the statement more optimistic (i.e. to ensure that “we get a fair share of the benefit, sooner, and broadly”):

Strong training and reskilling programmes targeted at people outside existing tech sectors, especially in regions with fewer tech firms.

Tax and regulatory policies that ensure profits generated from local infrastructure / services are taxed appropriately in the UK. Possibly local ownership or joint ventures.

Supporting local supply chains (equipment suppliers, maintenance, construction etc.) so more of the investment circulates through UK businesses.

Ensuring that data centres and infrastructure are spread geographically, so that the investment brings benefit to multiple regions, not only “hubs.”

Encouraging spillover / innovation: startups, smaller firms using the infrastructure; ensuring access to compute c

I will take the bait, four years time at the least, it isn’t exactly going to help out now in the short term when we need it. How many people will it employ? Not the tens or hundred of thousands that need work. Fine for technical expertise mostly if we have them, if not importing more people with given skill sets.

Then there is the long term profits, what percentage will we as a country get from the development and implementation around the world. The sceptic in me says that the profits will end up in the USA and not proportioned to here where the design and deve turning this board into a tabloid opinion column is another.

I then asked AI on the above and this is the reply

Your scepticism is entirely fair — the claim that Microsoft is investing “billions in the UK for AI” is true, but whether that translates into short-term benefit for most people, or fair share of profits/permanence, is another question. I’ll break down what we know, what the likely gaps are, and what the risks are.

What we know

These are some of the confirmed facts about Microsoft’s AI investments in the UK:

Microsoft is committing to invest £2.5 billion over the next 3 years to expand its next-generation AI datacentre infrastructure. Investing.com UK+3GOV.UK+3Computer Weekly+3

As part of that, Microsoft intends to bring more than 20,000 advanced GPUs to the UK by 2026. Investing.com UK+2Computer Weekly+2

It also has a program to train one million people in AI/digital skills. GOV.UK+2The Official Microsoft Blog+2

More recently, Microsoft has announced spending $30 billion in the UK by 2028 (i.e. including further AI infrastructure and operations). CNBC+1

So yes — the scale is very large, especially compared to past investments. There's infrastructure, compute capacity (GPUs, data centres etc.), and skills training.

What is unclear or less certain

Where reasonable doubts or caveats come in:

How many jobs will this create, especially in the short term

Microsoft has not (at least in all instances) given precise numbers of jobs created except broad statements like “exceptional individuals,” “skilled roles,” “training one million people” etc. Computer Weekly+3Microsoft UK Stories+3BBC+3

Many of the roles are likely to be highly technical (AI researchers, engineers, data centre operations, etc.). That means that unless someone already has relevant skills or can ‘upskill’ quickly, the barrier to entry will be high.

When will benefits be felt

Building data centres, installing infrastructure, setting up compute, training people — these take time. They don't immediately reduce unemployment or generate widespread income for people without technical skills.

The short-term economic benefit (say, within a year or two) is likely more evident in construction, firms supplying equipment, and in higher-paid technical roles, rather than mass employment or benefit across all sectors.

BB

“I like humans.”

You mean you like Starmer and Reeves?

ASI

“ AI-powered chatbots handle inquiries more efficiently, improving response times and customer satisfaction. satisfaction. In marketing, AI analyses customer data to identify trends and optimise campaigns, resulting in higher engagement and conversion rates. In finance, AI detects fraud and manages risk, enhancing security and building customer trust."

Well, it sounds the perfect solution to replacing Starmer and then if it is brilliant at finance it seems the best replacement for Reeves. AI takes out all personal feelings and views and funnily enough it can listen to customer satisfaction which is us and resolve the boat issue. It will be smarter in negotiations as well, a no brainer, how do we make it happen?

MV

“Aircraft seats booked but not used.”

Sounds like the Tories with flights to Rwanda and labour blocking them.

And it isn’t good news with the migrant situation with one in and one out. This will impact Lloyds because if they crucify Lloyds in November it is because they are paying indirectly for hotel fees in France now, so share price could go down and divis as well, see below.

The real plan is that the French are happy to put migrants in hotels but needed a justification to do it and then charge the UK at the same time. So now we are paying for a growing number of migrants to be housed in hotels on both sides of the Channel. This is part of Starmers master plan to support all migrants worldwide, don’t worry the UK can afford it, we will just tax the pensioners more and take their savings and inheritance from them until the pips squeak and they die from the cold in the winter because they can’t afford the heating bills.

HG

“ At least pensions will be up.. but of course those of you still working or earning will be paying more tax to pay for it.”

Don’t worry, it is all covered off by Starmer and Reeves with the growth plan kicking in. We shall all be better off, workers and pensioners alike.

Looking like the new state pension will be just shy of the personal allowance next year. The following year means all new state pensioners will officially be paying tax if they qualify for full amount. So any increase will actually be negated by some of it being taxed, which means the triple lock is falling behind again on supporters pensioners.

Do you mean the one where the AfD came in with 14.5% and the Greens with 13.5%.

The losers were the Greens, who seem to have lost a significant share of the vote, while the AfD almost tripled its result from five years ago, making them the winner in terms of percentage gains.

PDK

Your standard stock answer when you trot out a reply. When in a glass house avoid throwing stones.

Perhaps he wants to come clean about the stuff on the internet about Jenny Chapman and his alleged love child. Running a parallel with John Major, but love child not included.

Suddenly any squeaky clean image he wanted to keep has been blown clean out of the water. Then there is the dodgy deal on his CPS pension. Don’t do as I do, do as I say, so for a corrupt PM he is setting a great example on how the rest of us should behave. What a snake he is. Then there was Jenny, was she the person he cheated with while with his first wife? It only gets better.