Few VCs are experts in machine learning or building deep learning frameworks, but most of them are pretty good with unit economics.
Which is why they're laser-focused on generative AI's tech stack.
Whether it's infrastructure, middleware, applications or something else, investors are looking for founders who can dig defensible moats and dominate.
Full TechCrunch+ articles are only available to members.
Use discount code TCPLUSROUNDUP to save 20% off a one- or two-year subscription.
According to Leonard Wossnig, CTO of biopharmaceutical startup LabGenius, "the true value proposition of AI companies now lies not just within the models, but also predominantly in the underpinning datasets."
Due to "a noticeable lack of substantial differentiation," he says these models "are rapidly becoming commodities."
In this TC+ column, he presents questions that will help nontechnical VCs gauge a company's "data quality . . . and what could go wrong if the data’s not up to scratch," along with frameworks that show how each layer in the stack creates value.
Don't let the headline fool you: If you work inside an early-stage AI startup, you need to know which angles of attack investors are likely to take when probing your pitch for flaws.
Image Credits: E. Slomonson The Photo Group (opens in a new window) / Flickr (opens in a new window) under a CC BY 2.0 (opens in a new window) license.
TechCrunch Disrupt 2023 ended yesterday, and out of all the events I've attended since working here, this one was my favorite.
I moderated three panel discussions with investors, hosted a Q&A with TC+ columnist Sophie Alcorn, and I had the great pleasure of meeting and talking to scores of early-stage founders in the halls at Moscone Center.
We're all still catching our collective breath as my co-workers fly home to places like Pittsburgh, Paris and Providence, but keep an eye out next week for our recaps from Disrupt.
We uncovered a ton of actionable business intelligence and had some fun along the way.
Editorial Manager, TechCrunch+
How to spend your first $10K on paid ads
Image Credits: Westend61 (opens in a new window) / Getty Images
For his latest TC+ column, growth marketer Jonathan Martinez dug into a topic every startup faces at some point: How to allocate a budget for your first paid marketing campaign.
Starting with a hypothetical $10,000 budget, Martinez answers the following questions:
Which channels should a startup leverage?
How should a budget be created?
Which tests are the most important in the early stages?
“In the early days of your startup, it’s important to be as efficient as possible, not only with this $10K budget, but also with your time,” he writes.
Instead of fine-tuning an LLM as a first approach, try prompt architecting instead
Image Credits: Jorge Greuel (opens in a new window) / Getty Images
“Build versus buy” is a problem every startup faces, but in the era of generative AI, it can be an existential question.
Building an LLM from the ground up costs more than most companies want to pay, and modifying one to meet your specific needs requires costly data preparation.
“In contrast, prompt architecting involves leveraging existing LLMs without modifying the model itself or its training data,” says Victoria Albrecht, co-founder and CEO of Springbok AI.
“Instead, it combines a complex and cleverly engineered series of prompts to deliver consistent output.”
Ask Sophie: Can you explain the H-1B visa, EB-2 green card transfer and Visa Bulletin?
Image Credits: Bryce Durbin/TechCrunch
My startup is hiring. A leading candidate for one of the positions has an H-1B visa and has been waiting for an EB-2 green card for more than four years. This will be the first time our startup will navigate immigration.
Can you explain the H-1B visa and EB-2 green card transfer process? When do you stick with EB-2? The “Visa Bulletin” changed?
— Curious Co-Founder
How deep tech founders can secure early-stage fundraising in a downturn
Image Credits: Rosa María Fernández Rz (opens in a new window) / Getty Images
Software as a service is wide open, but deep tech founders hoping to connect with investors have a unique set of problems.
New technology often takes a long time to monetize, which means there’s a small community of VCs who have relevant interest and experience. During a downturn, that cohort can get even smaller.
Bucking the odds, French photonics company Cailabs acquired seed funding during a down market in 2013.
“Here we share the lessons we learned, which can help other deep tech founders looking for funding during a downturn,” writes founder/CEO Jean-François Morizur in TC+.
Get the TechCrunch+ Roundup newsletter in your inbox!
To receive the TechCrunch+ Roundup as an email each Tuesday and Friday, scroll down to find the “sign up for newsletters” section on this page, select “TechCrunch+ Roundup,” enter your email, and click “subscribe.”
8 web3 gaming experts discuss hurdles and opportunities in the road to wider adoption
Retro game console controller. Image Credits: mr_morton / Getty Images
Web3 gaming encourages players to earn NFTs and cryptocurrency and enables them to make collective decisions about the worlds they inhabit.
Investors and developers are expressing interest, but "web3 gaming still has some pretty difficult hurdles to overcome before it can go mainstream," writes Jacquelyn Melinek.
Don't use builds, reveals or animations in your pitch deck
Image Credits: Cavan Images (opens in a new window) / Getty Images
There’s a lot of pressure on founders to create pitch decks that will hold investors’ attention, but is it appropriate to use dynamic elements like animations in a presentation?
“Just don’t,” advises Haje Jan Kamps.
“In the thousands of decks I’ve seen over the years, I’ve never seen an animation truly add anything.”