If you've been turning to AI for answers lately, you might not trust it with your questions about AI.
That's where we come in.
Table of Contents
- What is AI?
- What can AI do?
- Will AI take my job?
- Are there AI ETFs you can invest in?
- Is AI bad for the environment?
- Can you date AI?
- Will AI replace Google Search?
- How can you tell if you're talking to AI?
- Can AI help you choose your investments?
- Does AI make you less intelligent?
- What if you don't want to use AI?
- Can AI save you money?
- What are the different types of AI?
- Is AI stealing?
- Who makes money from AI?
- Will there ever be Artificial General Intelligence (AGI)?
- Should you use AI?
- Can AI predict the stock market?
- Can AI be trusted?
- Is AI safe for kids?
What is AI?
AI stands for artificial intelligence. It's essentially a bunch of algorithms that program software and machines to make them imitate human functions such as thinking, creating, analysing, and replying.
Scientists and engineers develop the algorithms, and researchers study their impact.
What can AI do?
The dream is that AI will be able to replicate and then enhance anything humans can already do.
For the moment, AI can successfully:
- Perceive images and describe or categorise them
- Compare datasets to predict behaviour and make recommendations
- Create content such as writing, music, and video
- Translate human language
Will AI take my job?
It depends who you listen to.
CEOs of AI companies are convinced it can one day replace most workers.
Anthropic CEO Dario Amodei said that over the next one to five years, AI could replace half of all US based entry-level white collar jobs.
On the other hand, Meta is downsizing its AI department, while a recent MIT report stated that despite wide AI tool adoption, 95% of companies are currently getting zero return on investment.
All this is to say: the jury's still out. The consensus is that AI will have some sort of impact on most desk jobs, and especially for entry level jobs.
Are there AI ETFs you can invest in?
Are there ever. When you buy into an ETF, you get exposure to a selection of companies or other assets in one go.
AI ETFs can include the companies that make the chips that act as the brains of AI, the data centers that power the AIs themselves, and other companies that sell AI products, either in totality or as an add-on to existing product suites.
Is AI bad for the environment?
It's a mixed bag. On the surface, AI is really bad for the environment: it uses lots of energy and water to power data centers and keep them cool.
On the other hand, a 2024 Nature Communications study noted that AI could actually reduce building energy consumption by 8-19% by 2050, meaning they use less energy and produce less carbon pollution, which is huge because buildings account for about 40% of energy usage in the US.
Can you date AI?
If you want to. There are AI apps that provide virtual relationship experiences for users. There are also people who use generative AI such as ChatGPT in non-traditional ways. The My Boyfriend Is AI subreddit is filled with people who are sharing their experiences of AI relationships online.
Experts from the Mozilla Foundation warn that such products are 'on par with the worst categories of products' for privacy, however.
We think you should just remember to trust an AI as much as you would a normal internet stranger: don't give them your banking details or passwords, no matter how much of a sweet talker they are.
Will AI replace Google Search?
Google says no. What it is replacing is Google Search traffic for many websites.
Essentially, when Google introduced its AI answers feature, surfacing answers right at the top of the search results page, it reduced the need for browsers to click multiple sites to find further information.
It also pushed search results below the AI feature box, with one study showing up to 80% fewer clickthroughs for websites that had previously ranked highly, according to The Guardian.
It's an interesting downstream impact, because website traffic is usually what funds a website's existence itself, particularly for independent media and news.
How can you tell if you're talking to AI?
You can ask them how many bs are in blueberry. Just kidding.
It's hard to tell when you're talking to AI, but it turns out you can tell if you've been talking to AI.
One study by Cornell University showed that in the 18 months following the release of ChatGPT, humans increased their use of words such as 'delve', 'comprehend', 'boast', and 'meticulous'.
In fact, usage of the word 'delve' in podcasts and YouTube videos has increased 25-50% annually since ChatGPT.
The takeaway? If you want to protect your unique voice, diversify your information sources away from AI.
Can AI help you choose your investments?
I was mulling over how our VP of Investments, Jason Sedawie, says that you should think like an 'owner' instead of a 'consumer' when it comes to finance, and purchase stocks to own parts of great companies, rather than just buy what they sell.
In a nutshell: buy Starbucks or Popmart shares, instead of Grande lattes and Labubus, (if that's what you're into, not financial advice).
I was in line for coffee and thought, "I should really invest in coffee considering I drink so much of it," plus I've heard the beans are getting more expensive.
So I asked my AI: "What are some coffee ETFs?" and it answered me.
In essence: AI can help you choose your investments, but it can't guarantee you'll make money.
Someone from Spaceship (not in the investment team), who asked ChatGPT for trading strategies said, "Sometimes you're up, sometimes you're down! It's helped me discover some gems."
Another person noted, "For me, the deep learning function is a game changer, it's like having your own investment analyst."
(Keep in mind, you should always seek personal financial advice and not rely on AI solely to make your investment decisions.)
Does AI make you less intelligent?
It might make you feel more intelligent if you've never used words like delve or meticulous before.
Research shows that people who use it at work tend to use it in place of higher-order critical thinking skills, intervening only when they think the AI responses aren't measuring up. This TechCrunch article is an easier read than the Microsoft and Carnegie study it refers to.
An MIT study also showed that using your actual brain only to complete academic tasks creates stronger neural pathways than using a large language model, or even search engine assistance, does. Which kind of sounds like a… no brainer.
Trade offs abound. Some AI apps are enhancing education, such as teaching languages – and Duolingo did teach me how to ask for coffee before I went to Vietnam last year.
What if you don't want to use AI?
If you're anti-AI, you're likely opposed to using it to generate images, answers, or content for things you deem to be superfluous. That's the AI known as generative AI.
Ultimately, there are elements of AI in most of the ways we interact online: from Netflix's recommendation algorithm to the alerts your bank sends you if it thinks someone's stolen your card details.
Can AI save you money?
Potentially. I found thousands of dollars worth of potential savings when I asked AI to interrogate my spending, and again when I used it to help me plan a trip to Bali.
People have used AI to help them negotiate, write resumes, and apply for jobs. Some businesses have found the savings to be immense, depending on their industry.
What are the different types of AI?
Narrow AI: When a computer can perform a task as well as a human can, but the task is defined by the human. Examples include unlocking your phone with facial recognition; or categorising an email as spam.
General AI: AI that can match or beat human intelligence. Examples include robots that could think and act for themselves. To be clear, this doesn't exist, and may never, though proponents are sure it's just around the corner.
Artificial Super Intelligence (ASI): Basically an AI that's superior to humans in every way. This is entirely theoretical.
Is AI stealing?
AI models are trained on data, and that data isn't always given freely.
For example: Volpara Health is a company that uses AI to detect breast cancer. Its algorithm is made smarter by the agreements it has to learn from the 130 million mammograms it's taken of its clients.
On the other hand, OpenAI, Microsoft, Anthropic, and other companies have ingested large amounts of creative works arguing it's 'fair use'.
Copyright lawyers beg to differ, while Anthropic was subject to a class action from authors objecting to its use of their work.
Who makes money from AI?
Currently NVIDIA makes arguably the most money from AI. When the chip company became the first ever to reach $4 trillion in market cap, it was worth more than 97% of all world economies, and total global military spending, according to Forbes magazine.
The biggest beneficiaries of AI have been the hardware and software makers.
Will there ever be Artificial General Intelligence (AGI)?
Depends who you ask. Tech analyst Benedict Evans described the race toward AGI to the Guardian as being 'very vibes based'. Pretty much, nobody can say for sure.
Should you use AI?
Demand and salaries for jobs with AI skills keep growing, according to Bain, so if you're chasing a big pay packet it could be a good direction to head in.
Ultimately any new technology brings with it a series of ethical and moral questions that nobody can answer for you.
Personally, I'll never get the same joy out of using ChatGPT as I will from crafting a sentence but as a writer, that's to be expected.
Others might experience the joy of expressing themselves properly for the first time through AI assistance, and that's nothing to shun either.
Can AI predict the stock market?
Nah. Not yet, anyway. People have been using algorithms to trade since the invention of the computer.
If you've been around these parts before, you'll know that the markets are disrupted by everything from pandemics and trade wars to natural disasters and tweets.
Until AI can predict those things, it's up to us and our own hypotheses and strategies to give it our best.
Can AI be trusted?
We think you should trust AI as much as you would any stranger you meet online. Always verify the information it provides.
Protect your data.
Nick from our Engineering Team takes inspiration from Mr Weasley and says "Never trust something that can think for itself if you can't see where it keeps its brain."
Is AI safe for kids?
Another one for parents to decide. There are elements of AI that can be seriously shady including the unpredictability of its output.
The eSafety Commissioner warns away from any AI that simulates a personal relationship and has practical strategies for setting boundaries.
One of our engineers had this take:
"My two cents for at least LLM chat AI is that it's comparable to general internet access. It's both good and bad.
As long as the kids are taught how to use it properly, understand its bias and error and remain skeptical. It's an amazing tool, but that's all it is - it's not your friend, but it can be a good teacher."
