.svg)
BUILD DIFFERENT
How A.Team Helped McGraw Hill Build 'TikTok for Studying'
After more than a century in the textbook business, McGraw Hill decided to take on a seemingly impossible task: Create a direct-to-student app that would make studying fun.
At first they brought in a small three-person A.Team to build a prototype and test with students. Following an overwhelmingly positive response, the team grew to eight product builders as it built the MVP, and eventually to 27 buildersâspanning software architects, product managers, mobile developers, and growth marketersâleading up to the launch in October 2022.
The result is Sharpenâa study app that provides a continuous feed of bite-sized videos, swipeable study tools, and quizzes that mimics the feel of Instagram Reels with the gamified incentives of Duolingo.
Call it TikTok for Textbooks.
Itâs addictive. Itâs delightful. And in the business world, itâs the rarest thing of all: an example of digital transformation that actually worked. Taking on an established industry with a long history of predictability and rigidity is no easy feat.
McGraw Hill knew that the only way to move forward and scale was to find the right people. That's where A.Team came in.
âCHART OF THE WEEK
Are We Cruising for a Generative AI Bubble?
Brace yourself, we might be racing towards the next tech bubble! Now that Web3 has fizzled thereâs a hype void. So VCs are pouring money into Generative AI startupsâ$2.1 billion last year alone.
In October, Jasper AI raised $125 million in Series A funding. And now Microsoft is mulling another bet on OpenAIâto the tune of $10 billion, which would give it a $29 billion valuation.
But there are lots of reasons to be AI-skeptical. Like yesterdayâs bombshell report in Time on how OpenAI used outsourced Kenyan laborers earning less than $2 per hour to flag toxic content.
Or how hard it is to guarantee that the answers you get are truthful.
âThese systems have no conception of truth. Sometimes they land on it and sometimes they donât⊠Itâs just auto complete, and auto complete just gives you bullshit,â as Gary Marcus vividly explained on the Ezra Klein show.
Then thereâs the speculation that generative AI might replace Google. But the costs of running queries are still so high. When Morgan Stanley analysts compared ChatGPT's cost-per-query to Google's cost-per-search and found that OpenAI's costs are about seven times higher than Google's. The difference? The compute intensity of ChatGPT's natural language model is still too great.
That doesnât mean AI wonât be lucrative. And while valuations may appear frothy, companies are continuing to leverage more AI capabilities. Itâs always interesting to watch advanced primates figure out how to use a new toolâwho knows, the most useful application for AI might end up being something we never predicted.
On the brightâalbeit dystopianâside, Microsoft recently announced VALL-E, a new AI that can simulate anyoneâs voice from a 3-second sample. So even if you do spend all your money on AI, at least you wonât be broke and lonely.
MISSION MUST-READS
Build Different, Ep. 1: Apprentice and the COVID Countdown
As the global effort to distribute COVID vaccines ramped up in early 2021, Apprentice realized it'd need to build an entirely new version of its product to meet the needs of the worldâin just 45 days.
Why Fractional Teams Are the Secret to Hiring in 2023
Thereâs a new model for bringing in talent to accelerate product development while cutting fixed costs: fractional teams.
Freelancers Are Happier With Their Pay Than Everyone Else
Feeling underpaid? More Americans are finding financial satisfaction and professional autonomy in freelance work.
PARTING TWEET
.jpeg)
.jpg)
.jpg)
