AI-Generated Articles: Will AI Replace Human Writers?

Originally published on mathamagicians.co

Machines and humans have long since started working together. A single voice command can open the house lights, start a call, or even drive a car. It wasn’t a problem for any of us, albeit the initial fear of being “taken over by robots,” up until AI started to tap into the tasks we never imagined it crossing.

The arts have always been for human expression. We always believed that experiences shaped the things we could create. So when AI-created art entered the picture, it’s no surprise that many were not happy with the development.

AI and the writing industry

It’s scary to see news pop up of journalist layoffs and machine-generated articles getting published. There’s even been an ongoing debate on how AI can affect the writing industry, particularly journalism.

The million-dollar question is, “Will AI kill or save journalism?” Currently, newsrooms use AI for more tedious tasks, like gathering information and distributing articles. For work that involves deeper research and writing, we’re still dependent on human journalists.

However, media outlets are getting more experimental with using AI in their regular work tasks. The Wall Street Journal tried it out for transcription, while The New York Times tested AI for personalizing newsletters and moderating content. Indeed, a lot of functions have been replaced with technology, and journalists are encouraged to adapt to the changing times.

GPT-3 and the published article from The Guardian

In 2017, OpenAI sparked controversy after introducing GPT-2 – an AI program that could create articles from the very little info it’s fed. The technology was so advanced that even the company and its partner Elon Musk refused to make the code available to the public.

Today, we’re met with another breakthrough from the same company. Say hello to OpenAI’s new AI language generator, called GPT-3, which created a lot of buzz with its potential. Many were excited to try out the new program. Could it really write articles better than experienced writers?

Media outlet The Guardian was one of the first to try it out and recently released an op-ed written by GPT-3. The topic? To convince us, humans, that robots come in peace. The bot was fed a few requirements and a prompt for the piece before it started generating the article.

From the get-go, the article’s grammar is flawless. The sentence flow is impeccable. You wouldn’t be able to tell that it’s machine-written. All of these are impressive, except that AI is only a language generator. It has no touch of reality.

GPT-3 definitely has the potential to create compelling articles. Even the editors of The Guardian could admit that its output is a step ahead of man-written text. However, a quick experiment by The Technology Review shows that it still has a long way to go in terms of comprehension and reasoning.

Described as “arguably useful, but never reliable,” The Technology Review reveals that GPT-3 has yet to perfect its causal understanding and coherence. Its problem lies not with syntax, but with semantics. And this is because it only learns how people use words, not from real-life experiences like humans do.

Friend not foe

Everyone knows that AI is extremely powerful, and no one doubts what else it can do in the future. However, what makes writing unique is the thought that comes into the words, similar to how we value handmade items over mass-produced commodities.

If you think about it, machine articles aren’t that far off. It produces content from what we feed it. Instead of looking at it as a threat, perhaps we could further explore the possibility of creating better art through human experiences and machine abilities.

The fear of AI advancement (and the “impending doom” that it will someday wipe out humankind) stems from mere assumptions and a few horror movies. We tend to forget that it is we who feed these machines and dictate how it should respond to us.

We’ve come a long way in terms of AI, and we’re nowhere near creating the perfect bot. Until then, let’s embrace innovation, adapt to the new technology, and allow it to do what we intended it for in the first place: to make our lives easier.

Facebook’s new chatbot touches on Artificial Empathy; their “most human bot” yet

Originally published on mathamagicians.co

Conversations and understanding amongst humans require common sense and empathy – characteristics that set human exchange apart from machine-generated dialogue. With Siri or Alexa already exceeding expectations for humanized bots, Facebook is stepping in to create AI with a little more personality.

The tech giant’s recent development, a chatbot designed by its AI team, combines personality, general knowledge and a little artificial empathy to mimic human chitchat. Dubbed ‘Blender’, the new chatbot builds on new AI innovation and language picked up from the social platform. Among its potential uses in the future are functions that need to be more engaging and personal such as voice assistants and auto-complete algorithms.

While Facebook is yet to launch Blender fully, the company has so far shared snippets of the bot chatting with users online about Game of Thrones, vegetarianism, and a handful of other random topics.

Banking on actual conversations from sites like Reddit, Blender captures emphatic conversations and personalities to use for its training data.

“Scale is not enough,” says Facebook research engineer Emily Dinan (check out some of her amazing research on text, NLP, behaviour and conversational AI here), who worked alongside the team to create Blender. “You have to make sure you’re fine-tuning to give your model the appropriate conversational skills like empathy, personality, and knowledge.”

Baby Food For Thought: Preparing Bulgur For Your Child’s First Meal

Originally published on goodwill.market

One of the joys of being a parent is witnessing your little one’s exciting milestones, like when they have a bite of their first meal.

Entering parenthood is an exciting chapter in our lives, but it comes with its own challenges. For one, feeding your baby gets trickier as they get older. When they reach eight to nine months, it will only be a while before the messy eating phase begins. For first-time parents, nothing will test your creativity more than preparing their first solid food.

There are pre-made options readily available in the market, but if you’re looking for a more wholesome meal, homemade baby food might be a better option. A recent study from the University of Glasgow revealed that baby food from markets contains only half the energy and protein value of homemade meals. Unlike store-bought packs and jars that are thickened with preservatives, preparing their food in the kitchen allows you to choose only the freshest and healthiest ingredients.

For starters, bulgur makes an excellent base for baby food when you’re introducing them to solids. The consistency of whole-grain exposes them to different food textures and encourages them to use their mouths and tongues—an essential part of speech and feeding development. On top of that, the versatility of bulgur wheat makes it a darling to pair with fruits, salads, and other side dishes according to your little one’s liking.

Serving homemade meals is the first step to raising a happy, healthy eater.

If you’re keen on learning how to create meals to please their tiny palate, start off with a bulgur-based healthy meal. Here’s a quick and easy recipe for the next meal time with family.

What you’ll need:

• 1 cup Duru Fine Bulgur

• 3 bananas

• 100g celery

• 150g carrots

• 2 tbsp of honey

• 1 tsp of ginger

• 1 tsp of turmeric

Here’s how to do it:

1. Put the peeled and coarse chopped carrots and celery into a pan and add water to cover them.

2. Boil the veggies until they get very soft.

3. Pass the mix through a wire mesh screen to mash.

4. Shred the bananas.

5. Place bulgur in a bowl of hot water. Cover it and let the it absorb the water

6. Flavor it with honey, turmeric, and powder ginger. Serve as warm or cold. — (TGM)