Artificial intelligence (AI) is fast becoming the new “Marmite”. Like the salty spread that polarizes taste-buds, you either love AI or you hate it. To some, AI is miraculous, to others it’s threatening or scary. But one thing is for sure – AI is here to stay, so we had better get used to it.

In many respects, AI is very similar to other data-analytics solutions in that how it works depends on two things. One is the quality of the input data. The other is the integrity of the user to ensure that the outputs are fit for purpose.

Previously a niche tool for specialists, AI is now widely available for general-purpose use, in particular through Generative AI (GenAI) tools. Also known as Large Language Models (LLMs), they’re now widley available through, for example, OpenAI’s ChatGPT, Microsoft Co-pilot, Anthropic’s Claude, Adobe Firefly or Google Gemini.

GenAI has become possible thanks to the availability of vast quantities of digitized data and significant advances in computing power. Based on neural networks, this size of model would in fact have been impossible without these two fundamental ingredients.

GenAI is incredibly powerful when it comes to searching and summarizing large volumes of unstructured text. It exploits unfathomable amounts of data and is getting better all the time, offering users significant benefits in terms of efficiency and labour saving.

Many people now use it routinely for writing meeting minutes, composing letters and e-mails, and summarizing the content of multiple documents. AI can also tackle complex problems that would be difficult for humans to solve, such as climate modelling, drug discovery and protein-structure prediction.

I’d also like to give a shout out to tools such as Microsoft Live Captions and Google Translate, which help people from different locations and cultures to communicate. But like all shiny new things, AI comes with caveats, which we should bear in mind when using such tools.

User beware

LLMs, by their very nature, have been trained on historical data. They can’t therefore tell you exactly what may happen in the future, or indeed what may have happened since the model was originally trained. Models can also be constrained in their answers.

Take the Chinese AI app DeepSeek. When the BBC asked it what had happened at Tiananmen Square in Beijing on 4 June 1989 – when Chinese troops cracked down on protestors – the Chatbot’s answer was suppressed. Now, this is a very obvious piece of information control, but subtler instances of censorship will be harder to spot.

Trouble is, we can’t know all the nuances of the data that models have been trained on

We also need to be conscious of model bias. At least some of the training data will probably come from social media and public chat forums such as X, Facebook and Reddit. Trouble is, we can’t know all the nuances of the data that models have been trained on – or the inherent biases that may arise from this.

One example of unfair gender bias was when Amazon developed an AI recruiting tool. Based on 10 years’ worth of CVs – mostly from men – the tool was found to favour men. Thankfully, Amazon ditched it. But then there was Apple’s gender-biased credit-card algorithm that led to men being given higher credit limits than women of similar ratings.

Another problem with AI is that it sometimes acts as a black box, making it hard for us to understand how, why or on what grounds it arrived at a certain decision. Think about those online Captcha tests we have to take to when accessing online accounts. They often present us with a street scene and ask us to select those parts of the image containing a traffic light.

The tests are designed to distinguish between humans and computers or bots – the expectation being that AI can’t consistently recognize traffic lights. However, AI-based advanced driver assist systems (ADAS) presumably perform this function seamlessly on our roads. If not, surely drivers are being put at risk?

A colleague of mine, who drives an electric car that happens to share its name with a well-known physicist, confided that the ADAS in his car becomes unresponsive, especially when at traffic lights with filter arrows or multiple sets of traffic lights. So what exactly is going on with ADAS? Does anyone know?

Caution needed

My message when it comes to AI is simple: be careful what you ask for. Many GenAI applications will store user prompts and conversation histories and will likely use this data for training future models. Once you enter your data, there’s no guarantee it’ll ever be deleted. So  think carefully before sharing any personal data, such medical or financial information. It also pays to keep prompts non-specific (avoiding using your name or date of birth) so that they cannot be traced directly to you.

Democratization of AI is a great enabler and it’s easy for people to apply it without an in-depth understanding of what’s going on under the hood. But we should be checking AI-generated output before we use it to make important decisions and we should be careful of the personal information we divulge.

It’s easy to become complacent when we are not doing all the legwork. We are reminded under the terms of use that “AI can make mistakes”, but I wonder what will happen if models start consuming AI-generated erroneous data. Just as with other data-analytics problems, AI suffers from the old adage of “garbage in, garbage out”.

But sometimes I fear it’s even worse than that. We’ll need a collective vigilance to avoid AI being turned into “garbage in, garbage squared”.

The post Garbage in, garbage out: why the success of AI depends on good data appeared first on Physics World.

Dear Physics World readers, I’m going to let you in on a secret. I get anxious every time I see the word “networking” on a meeting or conference agenda. I’m nervous whether anyone will talk to me and – if they do – what I’ll say in reply. Will I end up stuck in a corner fiddling on my phone to make it seem like I want to join in but have something more important to do?

If you feel this way – or even if you don’t – please read on because I have some something important to say for anyone who attends or organizes scientific events.

Now, we all know there are many benefits to networking. It’s a good way to meet like-minded people, tell others about what you’re doing, and build a foundation for collaboration. Networking can also boost your professional and personal development – for example, by identifying new perspectives and challenges, finding a mentor, connecting with other organizations, or developing a tailor-made support system.

However, doing this effectively and efficiently is not necessarily easy. Networking can also soak up valuable time. It can create connections that lead nowhere. It can even be a hugely exploitative and one-sided affair where you find yourself under pressure to share personal and/or professional information that you didn’t intend to.

Top tips

Like most things in life, what you get from networking depends on what you put in. To make the most of such events, try to think about how others are feeling in the same situation. Chances are that they will be a bit nervous and apprehensive about opening the conversation. So there’s no harm in you going first.

A good opening gambit is to briefly introduce yourself, say who you are, where you work and what you do, and seek similar information from the other person. Preparing a short “elevator pitch” about yourself makes it easier to start a conversation and reduces the need to think on the spot. (Fun fact: elevator pitch gets its name from US inventor Elisha Otis, who needed a concise way of explaining his device to catch a plummeting elevator.)

Make an effort to remember other people’s names. I am not brilliant at this and have found that double checking and using people’s names in conversation is a good way to commit them to memory. Some advance preparation also helps. If possible, study the attendee list, so you know who else might be there and where they’re from. Be yourself and try to be an active listener – listen to what others are saying and ask thoughtful questions.

Don’t feel the need to stick with one person or group of people for the whole the time. Five minutes or so is polite and then you can move on and mingle further. Obviously, if you are making a good connection then it’s worth spending a bit more time. But if you are genuinely engaged, making plans to follow up post event should be straightforward.

Decide the best way to share your contact details. It could be an iPhone air drop, taking a photo of someone’s name badge, sending an e-mail, or swapping business cards (seems a bit unecological these days). If there are people you want to meet, don’t be afraid to seek them out. It’s always a nice compliment to approach someone and say: “Ah, I was hoping to speak to you today; I’ve heard a lot about you.”

On the flip side, avoid hanging out with your cronies, by which I mean colleagues from the same company or organization or people you already know well. Set yourself a challenge to meet people you’ve never met before. Remember few of us like being left out so try to involve others in a conversation. That’s especially true if someone’s listening but not getting the chance to speak; think of a question to bring that person into the discussion.

Of course, if someone you meet doesn’t seem to be relevant to you, don’t be afraid to admit it. I’m sure they won’t be offended if you don’t follow up after the meeting. And to those who are already comfortable with networking, remember not to hog all the limelight and to encourage others to participate.

A message to organizers

Let me end with a message to organizers, which – I’ll be honest – is the main reason I’m writing this article. I have recently attended conferences and events where the music is so loud that people, myself included, have gone with the smokers to the perishing cold outside simply so we can hear each other speak. Am I getting old or is this defeating the object of networking? Please, no more loud music!

I also urge event organizers to have places where people can connect, including tables and seating areas where you can put your plates and drinks down. There’s nothing worse than trying to talk while juggling cutlery to avoid a quiche collapsing down the front of your shirt. Buffets are always better than formal sit-down dinners as it provides more opportunity for people to mix. But remember that long queues for food can arise.

So what has networking ever done for me? Over the years the benefits have changed, but most recently I have met some great peer mentors, people whom I can share cross-industry experience and best practice with. And, if I hadn’t been at a certain Institute of Physics networking event last year and met Matin Durrani, the editor of Physics World, then I wouldn’t be writing this article for you today.

I’ll let you, though, be the judge of whether that was a success. [Editor’s note: it certainly was…]

The post Honor Powrie: ‘So what has networking ever done for me?’ appeared first on Physics World.