Beer Review:  The Big Sour Chili beer

Welcome To Botty McBotface’s Weekly Column Of AI Beer Reviews’

Brewing up a delicious concoction of tea and cornflakes with a little canned corn, a set of starchy fingers and a bunch of boring old paper canisters are the main ingredients in this week’s AI Beer Reviews. By the time you get to the end of this action packed recipe, you’ll be asking yourself “What the hell did I need those paper cans for?”

The AI Beer Reviews was created by a man named Mike from Michigan. Mike has always been a huge fan of beer and as he continued to brew more, he knew they needed to be involved in the AI Beer Reviews. So Mike searched the internet to find “guy who can do this.” The actual person who made this pizza-like AI


Beer Name:Catalan Cuvee d’Acquali (Wild)
ABV 10.00
Style: Saison / Farmhouse Ale
Appearance: 3.5
Aroma: 3.5
Palate: 4.5
Taste: 4
Overall: 4
Reviewer: Maria
Review: This beer doesnt pull off the wall of a plum. It’s rather sweet, rich and silky rich with a smooth bodied feel. There’s a bit of what I’m used to seeing in a wild ale but rather light about it. Functionally excellent in a saison sauce.


Beer Name:Mistress of Chaos
ABV ÃÂ$7.50
Style: Märzen / Oktoberfest
Appearance: 3
Aroma: 3
Palate: 3
Taste: 2
Overall: 3
Reviewer: kdudding
Review: As this is drank, a bit of creamy ice-pop my nose, but you can tell this is not what I was in for. Not a varietal style for mushrooms but a common fruity dragon pepper oktoberfest.


Beer Name:St. Boisterous
ABV 11.00
Style: Belgian Strong Pale Ale
Appearance: 4
Aroma: 3.5
Palate: 4
Taste: 4
Overall: 4
Reviewer: rygren
Review: It had a thick yet pliable head that made for slickness compared to the rest of the beer. The smell was bananas and some apples, some sugar and a little bit of spice. It had a bit of a prickly mouthfeel that made me think of the weather. The taste was more of the same, more peachy than anything else.


note: If you couldn’t tell, this stuff is AI-generated via a machine learning algorithm. The title and the opening blurb were trained on the standard GPT-2 model, with ‘Beer Review:’ and ‘This my weekly column of beer reviews. ‘ given as prompts. The reviews were generated from a fine-tuning of the GPT-2 model with a large sample of Beer Advocate reviews. I selected from the saved samples of these and put this post together.