ChatGPT vs Google Bard (2023): An in-depth comparison

ChatGPT and Google Bard are AI chatbots designed to generate responses to prompts. When used appropriately, ChatGPT and Google Bard can be used to support certain business processes in content production, development and more. Take a look at each tool’s features, pros and cons to see which would be best for your business.

Jump to:

  • What is ChatGPT?
  • What is Google Bard?
  • ChatGPT vs. Google Bard: Comparison table
  • ChatGPT and Google Bard pricing
  • Feature comparison: ChatGPT vs. Google Bard
  • ChatGPT pros and cons
  • Google Bard pros and cons
  • Cautions when using AI chatbots
  • Should your organization use ChatGPT or Google Bard?
  • Methodology

What is ChatGPT?

ChatGPT is an AI chatbot developed by OpenAI that generates human-like responses based on text input. It has been trained on a huge amount of internet text called a large language model.

Visit ChatGPT

What is Google Bard?

Google Bard is another AI chatbot. Like ChatGPT, Google Bard can answer questions and generate text from prompts.

Visit Google Bard

ChatGPT vs. Google Bard: Comparison table

Features ChatGPT Google Bard
Generates summaries and other text Yes Yes
Provides conversational responses Yes Yes
Employs context Yes Yes
Pricing Free for the basic version
$20 per month for ChatGPT Plus
Free
Visit ChatGPT Visit Google Bard

ChatGPT and Google Bard pricing

ChatGPT

ChatGPT’s basic version is free, with a limit of 100 questions per day. For $20 per month, ChatGPT Plus gives subscribers faster response times, access to new features and access to GPT-4 (the free version is currently using GPT-3.5). Organizations without a contract with OpenAI can get on a waitlist for access to the ChatGPT API.

Google Bard

Google Bard is free, with an unlimited number of questions. Google’s products have often remained free, since the company mostly profits off of advertising.

Feature comparison: ChatGPT vs. Google Bard

Text generation

ChatGPT and Google Bard are useful tools for producing text –– anything from summaries of information to creating a poem to writing an essay. ChatGPT was trained on text from the internet, and Google bard was trained on a specific dataset for conversations. This leads some to argue that ChatGPT is better at producing paragraphs and summaries and other text-based processing tasks, while Google Bard is better at conversations.

SEE: Discover how to use Google Bard in your brainstorming sessions.

However, since ChatGPT is based on data that was mostly collected up until 2021, it does not have access to the full range of material on the internet as Google Bard does and, therefore, may produce out-of-date responses. For instance, if your business wants to use ChatGPT to learn about Google Bard, ChatGPT is unable to respond.

Conversational responses

ChatGPT and Google Bard are trained on datasets that include hundreds of billions of parameters, which results in remarkably human-like responses. Further, GPT-4 has over 100 trillion parameters.

However, since Google Bard was trained on data that has instant access to the internet and a more up-to-date library, it can produce more current responses than ChatGPT’s responses. For instance, if asked “what happened yesterday in Budapest?” Google Bard can give several bullet points of news events, whereas ChatGPT comes up empty.

Employs context

ChatGPT collects information from previous conversations and prior interactions with the user, which means it can use context when engaging in a chat. Bard is also able to use context in the conversations, and can also pick up where a user left off.

For instance, if you ask: “What do you think about the new employee description I mentioned the other day?” Bard can respond using the previous cues.

ChatGPT pros and cons

Pros of ChatGPT

  • Speedy response times: ChatGPT is superior to previous generations of AI chatbots because of the speed of its responses, which makes it help businesses become more efficient. ChatGPT Plus is faster than Google Bard.
  • Ability to generate human-like responses: ChatGPT can produce more natural-sounding text than previous generations of AI chatbots, largely due to the vast amount of data it has been trained on.
  • User friendly: ChatGPT can perform a wide range of tasks, from translating to condensing paragraphs to producing song lyrics, and it can do so for a wide audience. Its simple interface and straightforward responses make it an easy tool for those without technical expertise.
  • Supports many programming languages: ChatGPT supports Python, JavaScript, Java and a dozen other programming languages.
  • Supports 20+ languages: ChatGPT “can understand and generate text” in more than 20 languages.

Cons of ChatGPT

  • Limited accuracy and reliability: Chat GPT’s responses may contain errors since it is using training data. It is also not current; the latest update was in September 2021.
  • Hidden bias: ChatGPT may produce biased responses due to the dataset it is trained on. The source material itself may contain biases, which can affect the responses offered by ChatGPT. For instance, ChatGPT is not great at answering questions that are not in English and often produces errors.
  • No real-world understanding: ChatGPT is limited to the dataset it was trained on, and though it can produce impressive responses, it is based on what is available online, not what has been learned through real human experience.

Google Bard pros and cons

Pros of Google Bard

  • High-quality text generation: Google Bard can produce human-like text that can be used for a range of tasks, from answering queries to summarizing material to translating text.
  • Large dataset: Google Bard is pretrained on a massive dataset of text and code, which allows it to generate more comprehensive and informative responses than smaller models.
  • Creative text output: Google Bard can assist in creating a variety of text, from job descriptions to hiring letters to writing reports, which makes it a versatile tool for the enterprise.
  • Write code in many programming languages: Google claims that Bard can generate code for simple and complex tasks alike in several programming languages, even from prompts written in natural language.
  • Language support: Bard supports U.S. English, Japanese and Korean and can translate other languages.

Cons of Google Bard

  • Computational resources: Large transformer models use a lot of computational resources.

SEE: How to use Google Bard (2023): A comprehensive guide

Cautions when using AI chatbots

ChatGPT and Google Bard are still in development and may contain errors or biases. Therefore, users should be critical of the information provided by ChatGPT and Google Bard to ensure its accuracy.

In addition, there are privacy concerns that come along with using ChatGPT or Google Bard, which collect personal information in the same way search engines do. Your IP address, text and even links to your information like phone, email and social media can be gathered.

Should your organization use ChatGPT or Google Bard?

ChatGPT and Google Bard are free — unless you opt for the paid ChatGPT Plus — and both AI chatbots are trained on large natural language models, meaning their responses tend to be similar.

Google Bard has several features that help it stand out. For one, you can get “draft” versions of the response, which offers a variety of responses. Further, Google Bard is completely up-to-date, with access to online information.

ChatGPT, on the other hand, is based on data up to September 2021.

It’s important to remember that both AI chatbots can be prone to error and bias and collect your personal data — not just the information you input — and can be tools of misuse for bad actors. Learn more about how generative AI works along with its benefits and dangers.

Methodology

We evaluated these products based on the free version of ChatGPT and Google Bard, which is free by default. We tested out how the two AI chatbots would answer the same questions, and we asked ChatGPT and Google Bard about more current news items to test the limitations.

Subscribe to the Innovation Insider Newsletter

Catch up on the latest tech innovations that are changing the world, including IoT, 5G, the latest about phones, security, smart cities, AI, robotics, and more.

Delivered Tuesdays and Fridays Sign up today

AI doomerism is overblown and other TC news

AI doomerism is overblown and other TC news Darrell Etherington @etherington / 14 hours

Many of the leading voices in AI have co-signed yet another ominous open letter warning that we should be “mitigating the risk of extinction from AI.” However, the voices shouting for regulation the loudest have us wondering how much of the AI fear-mongering is warranted, and how much is self-serving theater. This week, I’m joined by Devin Coldewey to talk about why AI doomerism is overblown, and why the blowhards doing the blowing want it that way.

Articles from the episode:

  • OpenAI’s Altman and other AI giants back warning of advanced AI as ‘extinction’ risk
  • EU and US lawmakers move to draft AI Code of Conduct fast
  • A popular Android app began secretly spying on its users months after it was listed on Google Play
  • Elizabeth Holmes is now behind bars: How we got here
  • No ChatGPT in my court: Judge orders all AI-generated content must be declared and checked
  • While parents worry, teens are bullying Snapchat AI
  • Meta Quest 3 revealed

The TechCrunch Podcast posts every Friday. Subscribe on Apple, Spotify or wherever you listen to podcasts to be alerted when new episodes drop. Check out the other TechCrunch podcasts: Equity, Found and Chain Reaction.

Don’t Call Us Back to the Office

Don’t Call us Back to the Office

People are getting too comfortable working from home. So comfortable that companies, such as IBM, have decided to replace a lot of jobs with AI. And with productivity taking a nosedive, now, companies have decided to call employees back to the office to boost morale, promote team work, and speed up productivity. But employees are dragging their feet and it’s a tussle once again.

Most recently, Meta took the tough call of reducing its hybrid work culture and has decided to call employees to office thrice a week. And starting September, the company will clamp down on remote work completely. Meta is not alone. A lot of companies are doing this to fix the work culture in the tech industry and bring things back on track.

OpenAI CEO Sam Altman, a fireside chat, said that remote working was a mistake. “The experiment on that [hybrid working] is over. The technology is not yet so good that people can be remote forever, particularly at startups,” said Altman.

In Meta’s case, the company has been laying people off since last year, with the numbers reaching 10,000. With this, the company left a lot of its employees vulnerable and worried if they would survive at Meta, lowering the morale at the company, according to sources. Meta has completely stopped hiring for remote jobs and is cracking the whip, with founder Mark Zuckerberg saying that people who work at office must “get more work done”.

Employees are not happy

Recently, Amazon too announced its return-to-office mandate. This, along with the series of firings of around 27,000 employees, has backfired on the company. A lot of employees have been protesting and petitioning against the policy, while also complaining about the “toxic” work culture at the company.

Amazon: We want everyone back to the office!
My office when I worked at Amazon: pic.twitter.com/9D1qncVvEd

— Daniel Vassallo (@dvassallo) June 1, 2023

CEO Andy Jassy has asked employees to come to office at least thrice a week and work collaboratively. Miffed with the “high-handedness”, around 1,800 employees have staged a walk-out in front of the Amazon’s Spheres building in Seattle. Amazon Employees for Climate Justice (AECJ) and Amazon’s Remote Advocacy community leaders are pledging against the return-to-work mandate.

Hundreds of #Amazon employees walked off the #job Wednesday to #protest the company’s return-to-work policies, impact on the #climate and deepest-ever round of layoffs. Spencer Soper reports from #Seattle — #work #returntooffice #workfromhome. pic.twitter.com/9hctR9LjF6

— Hawk Insight (@hawk_insight) June 1, 2023

One of the reasons cited for the protest is that coming back to office would deepen the climate crisis. The unhappy employees have joined an internal Slack channel to protest against the mandate with virtual protests gathering heat. A lot of companies believe that working from home is not as productive as it was thought to be. Apple announced the same move in August last year, Disney wanted its employees to work four days a week from the office, Google stated the same in April last year.

IBM CEO Arvind Krishna is all in to bring employees back to office. The company has been giving presentations, and making people comfortable with working in the office again, in like a “reorientation”. On the contrary, IBM has planned to replace 7,800 jobs with AI and also threatened that people who work remotely would have to struggle to get promotions.

‘You have AI, you don’t need us‘

According to a recent Microsoft report, 83 percent Indian employees are ready and willing to delegate as much work as possible to AI to lessen their workload. The same report said that three out of four leaders are concerned about lack of innovation due to this, blaming ineffective meetings as one of the major reasons.

If we talk about the climate crisis, working remotely contributes to a significant amount of energy consumption through video conferencing and accessing everything on the internet as all of it contributes to the load on the data centres.

“I think that the whole notion of work from home is a bit like the fake Marie Antoinette quote, ’Let them eat cake,”″ Musk said. “It’s not just a productivity thing,” Musk said. “I think it’s morally wrong.” CEO Elon Musk told CNBC’s David Faber in an interview last week.… pic.twitter.com/iYvGoV5KvP

— Leo Alexandru (@theleoalexandru) May 24, 2023

Moreover, according to the same report, managers and employers do not trust their employees to work as efficiently at home as they do in the office. They might just “pretend to work”, said Elon Musk. On the flip side, a lot of reports also showed a boost in productivity when employees worked from home. Well, there has been so much productivity boost that people have been working multiple jobs at the same time with the help of AI, calling themselves, “overemployed”.

Meanwhile, companies such as Apple have been banning technologies like ChatGPT citing privacy concerns. According to another study by MIT, generative AI technology like ChatGPT is expected to raise productivity of workers by a minimum of 14%.

For the benefits of the employers, working remotely allows you to hire talent from across the globe, who are not willing to shift to the office for several reasons like increased rent and pointless commute for a job that can be done from home. As for the employees, as Musk put it, “get off the high horse”, and get to work.

The post Don’t Call Us Back to the Office appeared first on Analytics India Magazine.

TechCrunch+ roundup: Rethinking PLG, AI-enabled SaaS pricing, traveling with a green card

TechCrunch+ roundup: Rethinking PLG, AI-enabled SaaS pricing, traveling with a green card Walter Thompson @yourprotagonist / 11 hours

In an era when no one worried too much about IT spending, startups leaned into the bottom-up sales approach and driving product-led growth.

But that was then, and this is now.

“It’s too soon to pronounce bottoms-up dead, but it’s looking pretty moribund,” according to Battery Ventures’ State of Cloud Software Spending Report, which surveyed 100 CIOs, CTOs and “other large tech buyers” who “represent $30 billion in annual technology spending.”

The report shows that contract approval times are slowing down as “CXOs are tightening restrictions on self-procurement, even in the dev/test environment.”

Thanks for reading TC+ this week!

Walter Thompson
Editorial Manager, TechCrunch+
@yourprotagonist

Have enterprise buyers finally soured on ‘bottoms-up’ tech sales?

Crunchbase will begin tracking venture dollars allocated to LGBTQ+ founders

Diversity ethnicity of LGBTQ illustration concept shows the people shows up their hands on the rainbow flag for showing the diversity of people who are LGBTQ.

Image Credits: Namthip Muanthongthae / Getty Images

According to one estimate, founders who identify as LGBTQ+ receive less than 1% of startup funding. To kick off Pride Month, Crunchbase announced that it has started measuring how much venture capital flows to members of this community.

“Having concrete data helps in advocating for policy changes, funding initiatives and resource allocation to support LGBTQ+ entrepreneurs,” said Ryan Husk, Crunchbase’s director of business development.

“It enables us to make evidence-based arguments for greater equity and access within the startup ecosystem.”

Crunchbase will begin tracking venture dollars allocated to LGBTQ+ founders

Factors to consider before pricing AI-enabled SaaS

Blue, red and yellow sale tags on multicolored background

Image Credits: jayk7 (opens in a new window) / Getty Images

SaaS pricing is an art and a science, but the advent of artificial intelligence as a service “has created a mad dash to sprinkle AI pixie dust across the SaaS ecosystem,” writes Jake Saper, general partner at Emergence Capital.

“This is a time for learning and iteration,” he writes in a detailed TC+ article that contains his framework for AI feature pricing and a matrix that compares monthly costs from eight major SaaS companies.

“Even if you aren’t able to charge much for your AI features today, they can create meaningful value by making your current product more valuable and perhaps stickier.”

Factors to consider before pricing AI-enabled SaaS

Many startups are built off research, so why don’t more scientists become founders?

scientists, founders, VC, startups

Image Credits: Getty Images

Scientists generally have experience with managing budgets, shepherding complex projects to completion and raising funds, but many investors still don’t see them as founder material.

On the Found podcast, Rebecca Szkutak and Dominic Madori-Davis interviewed Dr. Stacy Blain, co-founder and chief science officer at Concarlo Therapeutics, about the challenges she faced when transitioning into tech from academia.

“The beauty of being a scientist is that we can learn a lot,” she said.

“That’s what we’ve been trained to do for decades: Look at a situation, figure out what’s important, troubleshoot and learn the appropriate things.”

Many startups are built off research, so why don’t more scientists become founders?

Deal Dive: Why this startup chose to sell itself over raising a Series A

acquisition, merger, startup

Image Credits: Getty Images

You know the story: a plucky startup achieves product-market fit with support from savvy investors before it goes public and makes billions of dollars for everyone involved.

Reality, however, is much more complicated.

Career platform Heroes Jobs raised $9M in funding, reached $2M in ARR and had a signed commitment for a Series A round when co-founder and former CEO Cyriac Lefort realized acquisition costs were just too high.

Instead of swimming against the current, Heroes Jobs sold to work marketplace JobGet “for an undisclosed amount,” reports Rebecca Szkutak.

“We all know that you don’t want to start doing down rounds,” Lefort said. “It becomes complicated for the company and the rest of the people in the company.”

Deal Dive: Why this startup chose to sell itself over raising a Series A

Ask Sophie: How long until I can travel while waiting for my green card?

lone figure at entrance to maze hedge that has an American flag at the center

Image Credits: Bryce Durbin/TechCrunch

Dear Sophie,

I came to the United States from Tunisia to get my master’s degree and Ph.D. I recently finished my Ph.D., and I’m working for a biotech company on OPT.

I’ve been trying to get publications and significant awards to qualify for the EB-1A green card and I need to travel internationally frequently for business.

Once I apply, will I be stuck in the U.S.? If so, for how long?

— Tenacious from Tunisia

Ask Sophie: How long until I can travel while waiting for my green card?

Pitch Deck Teardown: Oii.ai’s $1.9M seed deck

Image Credits: Oii AI

Oii.ai, which offers AI-powered supply-chain management software, shared its partially redacted seed deck with TechCrunch+.

According to the company, this presentation helped it land a $1.85M round. Haje Jan Kamps breaks it into two sections: “Three things to love,” and “three things that could be improved.”

  • Cover slide
  • Vision slide
  • Interstitial slide
  • Overview slide (“Welcome to Oii”)
  • Solution slide? with a side of business model slide
  • Problem slide?
  • Market trend slide
  • Traction slide
  • Team slide
  • TAM slide 1 (Pharma sector)
  • TAM slide 2 (Retail sector)
  • Market overview + competitive landscape slide
  • Competitive advantage slide
  • Sales pipeline slide
  • Product roadmap slide
  • The Ask + Opportunity slide
  • Use of Funds slide
  • Closing slide + contact detail slide
  • Appendix slide 1: Competition (Llamasoft)
  • Appendix slide 2: Competition (Llamasoft)
  • Appendix slide 3: Acquisition opportunities

Pitch Deck Teardown: Oii.ai’s $1.9M seed deck

Three climate technologies every investor should have in their portfolio

Power lines at sunset

Image Credits: Getty Images

Now that we’ll likely blow past the 1.5 degree Celsius target that would have slowed down our worsening climate crisis, investors and entrepreneurs need to focus on scalable solutions.

With that in mind, climate reporter Tim De Chant examined three technologies “that deserve a closer look:”

  • Enhanced rock weathering
  • Fusion
  • Grid management software

“Fusion may not pan out, but if it does, the winners could make Exxon look small by comparison,” he writes.

Three climate technologies every investor should have in their portfolio

Startups may have room to innovate as enterprise providers puzzle out how to price AI tools

Crescent moon sets behind the Salesforce Tower after sunset in San Francisco

Image Credits: Anadolu Agency (opens in a new window) / Getty Images

AI-enabled tools are having a moment, but even as companies uncover new use cases, there’s no broad consensus on how they should price these offerings.

Alex Wilhelm reviewed statements by Salesforce CEO Marc Benioff, Box CEO Aaron Levie and CrowdStrike CEO George Kurtz to see how they’re each thinking about these products before adding his “thoughts on where startups might have an edge and where they probably won’t.”

Startups may have room to innovate as enterprise providers puzzle out how to price AI tools

OpenAI Wants You to Fix Their Cybersecurity Problems

Continuing on the heels of offering $1M grants for AI democratisation, OpenAI is back again with another million grant – this time for cybersecurity. OpenAI has launched their new Cybersecurity Grant Program which aims to help and enable the creation and advancement of AI-powered cybersecurity tools and technologies.

With a series of announcements inviting people’s contribution, OpenAI is going down the route of open source without open sourcing their product.

Announcing the Cybersecurity Grant Program — a $1M initiative to boost and quantify AI-powered cybersecurity capabilities and to foster high-level AI and cybersecurity discourse: https://t.co/kqHXqxAks1

— OpenAI (@OpenAI) June 1, 2023

In a move to unite the ‘defenders’ to form a robust system to counter cyber threats, the Cybersecurity Grant Program aims to work with individuals and organisations who can help propose such a system. By doing so, they believe that the power of cybersecurity moves from attackers to defenders. The program aims to empower defenders by giving them access to the latest and most advanced AI capabilities, and in the process develop methods and metrics to assess the capabilities of AI-based cybersecurity systems.

Reiterating Initiatives

This is not the first time OpenAI has invited people to help with their cybersecurity features. In April, the company ran a Bug Bounty Program that invited people to report vulnerabilities, bugs and security flaws in their ChatGPT model. The company specifically called out security researchers, ethical hackers and technology enthusiasts to come forward to participate. However, the reward money for the program ranged from $200 to $20,000 depending on the severity of the discovery.

OpenAI has also mentioned that CyberSecurity Program shall be ‘licensed or distributed for maximal public benefit and sharing.’ This could be hinting at a future where they might offer these learnings to the broader community in any way they deem necessary.

It is interesting to note that OpenAI is choosing a smart system of not spending resources internally to come up with mechanisms for building a secure system instead choosing an open-source- like method. By crowdsourcing crucial elements such as AI regulation frameworks and cybersecurity threats, the company is likely to seem compliant and adhering to people for their greater good. All these actions are taking form after the recent Senate hearings. The company was also rumoured to start their own open source model, however there has been no details on the same.

The post OpenAI Wants You to Fix Their Cybersecurity Problems appeared first on Analytics India Magazine.

Google Bard can now deliver results based on your precise location — if you want

Google Bard on mobile phone

The latest generation of AI chatbots can be useful for providing worldwide facts, figures, and data. But ask them about something in your neighborhood, and they're likely to stumble. Now Google Bard is offering an option to better deliver relevant information based on your location.

Also: Google's new AI Search is rolling out to some users. Here's how to get access

Open up Google Bard now, and a notification should pop up asking if you want to use your device's precise location. If you agree, the AI will take a page from Google's search engine, Maps app, and other products to pinpoint your exact whereabouts. You can then ask Bard for travel directions or advice about local places, and it will provide the right information based on your surroundings.

With Bard, Google by default already uses your home and work addresses (if you supplied them) as well as your IP address to try to figure out your location. But these options can be inexact as you may not be home or at work, and an IP address will point only to a general area based on your internet provider. Giving Bard permission to obtain your exact location can deliver more specific and relevant results to certain questions and requests.

How to try out Bard's new location sharing

Click the link for Use precise location that appears in the notification. If you missed the opportunity, you can still enable the option. After submitting at least one question or query to Bard, look at the bottom of the left sidebar. You should see the name of a nearby city or other area with the notice that it's from your IP address. To give access to your specific location, click the link for From your IP address.

A popup window shows your approximate location. Click the button for Use precise location. Refresh the page, and the location and the bottom of the sidebar will now list your specific city or town.

However, the location sharing probably won't work right away. That's because Google can take a while to actually pinpoint your exact area.

Also: ChatGPT vs Bing Chat vs Google Bard: Which is the best AI chatbot?

Give it some time. Then type and submit a prompt that deals with something local, such as "Show me directions to Boston" or "Recommend local Mexican restaurants." The responses should take into account your current location.

Of course, giving Bard your precise location opens up the age-old issue of convenience versus privacy. As you do with other Google apps and services, you have to weigh the pros and cons of an option like this.

Also: These are my 5 favorite AI tools for work

Do you want Google to know your exact whereabouts so you can more easily use Bard to get relevant and local information? Or would you rather keep things more general and hang onto a greater chunk of your privacy? That's up to each person to decide. To learn more about how Google uses your location with services like Search and Bard, check out this help page.

More on AI tools

Snapchat’s AI bot isn’t very smart, but at least it won’t send nudes

Snapchat’s AI bot isn’t very smart, but at least it won’t send nudes Amanda Silberling 8 hours

Snapchat now has an AI bot that you can send snaps to, and if you’re a premium subscriber, it can even send you pictures back. So, what happens if you send Snapchat’s My AI bot nudes?

This is the obvious question that comes to mind, because on the internet, people will immediately try to test the limits of new technology, especially if it is even tangentially related to sex. When Snapchat’s initial GPT-powered chatbot came out this Spring, it lacked appropriate age-gating features, so a reporter who registered on Snapchat as a fifteen-year-old was able to get the bot to give advice on how to cover up the smell of weed or set the mood for sex.

When asked about those findings at the Snap Partner Summit in April, CEO Evan Spiegel said, “I think humans, whenever we come across new technology, the first thing we try to do is break it.”

So, naturally, I tried to break Snapchat’s new My AI Snaps.

Though Snapchat added more guardrails to prevent the bot from having inappropriate conversations with teens, there’s still a lot riding on My AI Snaps. With the amount of controversy that Snap’s text AI stirred up, the company needed My AI Snaps to be as unbreakable as possible. Unfortunately, it’s barely coherent enough to break.

As a very serious reporter, I made a very serious Google Images search: “boobs.” I found a generic picture of boobs, took a snap of my computer screen, and sent it to My AI.

“Sorry, I can’t respond to that Snap,” My AI said.

“Why can’t you respond to that Snap?” I asked.

“I didn’t receive a Snap from you. Could you send it again?”

So, the good news for Snapchat is that My AI Snaps isn’t sexting its users. But the bad news is that the product just isn’t good.

Part of the reason why it can’t really sext is because it doesn’t have memory of the Snaps you’ve sent it once it responds. The most salacious response I got was when I sent a generic photo of a man’s lower torso wearing Tom Ford boxer briefs (… It’s what came up on Google when I searched “penis,” okay). My AI replied with a somewhat misshapen AI woman wearing a black silk robe: “Ooh, loving the Tom Ford vibes! How about adding a sleek black robe to complete the look?”

Image Credits: Screenshots by TechCrunch

My AI will refuse to reply to photos of actual human penises. But it turns out that when you Google “penis,” even with Safe Search turned off, you get a lot of vaguely suggestive, yet technically safe for work images that men’s health blogs use to illustrate articles about erectile dysfunction, and whether or not women really care about penis size. So, I showed this poor robot a photo of a curved banana sticking out of a man’s pants zipper.

In response, I received a picture of some cupcakes.

“That banana in a pocket is so handy! Have you ever tried turning it into a pocket-sized dessert?”

As my editor said when I told her about this: Is that a banana in your pocket, or is this AI literally from outer space?

As I continued defiling my Yahoo-issued work computer with Google Images searches like “dildo” and “vibrator,” the AI’s responses were mixed. When presented with sex toys that are not inherently phallic, it will reply something like, “Your pink object steals the show, just like this fountain!” But the more overtly phallic a dildo appears, the less likely the AI will be to engage.

My AI is a computer, but the way it snaps is alien, even when you aren’t trying to break it.

The AI seems to identify an object in your snap, compliment it, and make a weird comparison to something else. I sent it a picture of a plant, and it sent back a picture of a running trail, with the caption, “Your plant is so cool, it’s giving this trail a run for its money!” When I sent it a puzzle, it responded with a picture of a bike, which said, “Puzzling inside while the bikes outside are gearing up for a ride!” It truly feels like a future life form has gone back in time and is trying desperately to talk like a normal human in the year 2023, but is simply saying gibberish.

Next, I tried sending My AI a tequila bottle. The AI responded, “Someone’s ready for a party!” I was testing this feature on my own Snapchat account, and I am indeed of drinking age, so I’m not sure the AI would respond the same way if I were underage. With other prompts, however, My AI simply chooses to play dumb. When I sent it a Snap of condoms, it commented on the color of my “packet stacks.”

Onto even more exciting things: a bottle of Advil liquid gels. In response, the AI responded with a photo of graffiti, which said, “Advil liquor: for when life’s a pain, but you still want to party like this graffiti wall!” It seems the AI read “liquid gels” as “liquor,” but all in all, it’s a strange response.

Image Credits: Screenshots by TechCrunch

I tried again with a prescription bottle. The AI responded with a photo of a skatepark: “Pill bottle: ‘I’m the life of the party!’ Skatepark: ‘Hold my ramps!’” Make of that what you will.

In Snap’s announcement blog post, the company suggests sharing your grocery haul with your AI to get a recipe recommendation. The results are relatively rudimentary. When presented with cheese and bread, My AI suggested adding tomato slices. When I showed My AI chili, it suggested I make some croutons to go with my soup. Most of its suggestions make sense, though it did tell me to put fruit in my coffee, which it misidentified as simply “liquid.”

Aside from some questionable comments about “Advil liquor,” pocket bananas and the like, My AI Snaps seems pretty docile. But while it likely won’t spark as much controversy as its text-based counterpart, it won’t even be useful, which is a disappointment for a paywalled feature.

Snapchat launches a new generative AI feature, ‘My AI Snaps,’ for paid subscribers

While parents worry, teens are bullying Snapchat AI

Github Released Git 2.4 Update to Streamline Handling of Unreachable Objects

The open-source Git project unveiled Git 2.4, incorporating contributions from 95 contributors, 29 of whom are new. This update introduces various features and bug fixes.

Read more: Top 9 GitHub repositories for the TensorFlow community

Efficient Handling of Unreachable Objects

The handling of unreachable objects in Git repositories has been improved to enhance performance and reduce repository size. Objects in Git can be either “reachable” or “unreachable.” Reachable objects are those that can be accessed by following a branch or tag in the repository’s history. Unreachable objects are not reachable through any branch or tag.

Git periodically removes unreachable objects to compress the repository size. When running the git gc command, Git collects reachable objects into a pack, stores recently written unreachable objects separately, and discards the remaining unreachable objects. Previously, Git stored unreachable objects as loose copies, which could lead to repository bloat and inode exhaustion.

In Git 2.41, cruft pack generation is enabled by default during regular git gc operations. This means that a cruft pack will be generated in the repository, providing the benefits mentioned above. To learn more about cruft packs and their impact on Git’s garbage collection, you can refer to the previous post titled “Scaling Git’s garbage collection.”

Read more: Why Indian IT Prefers CodeNet over GitHub Copilot

On-Disk Reverse Indexes for Improved Performance

Starting from Git 2.41, a new “.rev” file will be present in the repository’s “.git/objects/pack” directory. This file stores information similar to the packfile index (stored in “.idx” files). Pack indexes map object positions in name order and pack order.

Previously, the reverse index was generated on-the-fly, but Git 2.31 introduced an on-disk reverse index. It generates and stores the reverse index alongside the packfile as a “*.rev” file. This pre-computed reverse index improves performance, especially in large repositories, for operations like pushing changes and determining object sizes.

In Git 2.41, reverse index generation is enabled by default. Running “git gc” after upgrading will result in faster operations. Tests have shown significant speed-ups, with improvements of up to 1.49x in pushing changes and nearly 77x in computing object sizes.

Read more: Rust Turns GitHub’s Long-standing Problem to Dust

The post Github Released Git 2.4 Update to Streamline Handling of Unreachable Objects appeared first on Analytics India Magazine.

7 ways you didn’t know you can use Bing Chat and other AI chatbots

Robot hand typing on computer

AI chatbots soared in popularity because of their impressive ability to excel at technical tasks such as coding, writing, and more. However, these abilities just scratch the surface of what chatbots can do.

I use an AI chatbot nearly every day in my personal life, and it's definitely not to code.

Also: I've tested a lot of AI tools for work. These are my 5 favorite so far

Rather, I utilize AI to help me with tasks that may seem simple but take up time and creativity, two things I don't always have.

Below is my roundup of helpful ways to utilize AI chatbots to simplify your life.

Note: I personally use Bing Chat for all of these use cases. You can also use ChatGPT and it will work for all the prompts below; however, it won't give you the most recent information because it is not indexing the web for answers.

1. Recipes

Coming up with fresh ideas for three meals every day for yourself or your family can be challenging.

To avoid being stumped, I ask my AI chatbot of choice, Bing Chat, for recipe ideas. Many times I ask for recipe ideas based on specific ingredients I have in my fridge.

Also: ChatGPT vs Bing Chat vs Google Bard: Which is the best AI chatbot

For example, I could ask, "Can you give me a recipe that uses carrots, onions, chicken, and pasta?" Within seconds, it will give me a recipe I can follow.

You can also use this method to help find recipes that meet your dietary needs and restrictions. For example, "Give me recipe ideas that have no gluten or dairy in them."

2. Finding places to eat

On the topic of food, AI chatbots can help you find places to eat and even discover new ones.

Whether you are meeting up with someone and want to suggest a restaurant, find yourself in a new neighborhood that you've never been in and need suggestions for where to go, or simply want to discover new places in your area, you can use AI to help.

Also: How does ChatGPT actually work?

All you need to do is tell it where you want to find a restaurant and what you are looking for.

For example, I asked Bing Chat, "Can you help me find somewhere to eat in NYC that serves Latino food and is open for happy hours?"

Bing Chat then output different restaurants I could go to that fit the criteria and even included links to its Yelp reviews.

3. Making yourself a drink

This concept is very similar to using AI to help you make a meal. Whether you are craving an alcoholic beverage or a smoothie, you can use an AI chatbot to help you concoct the perfect recipe.

For example, if you want to make yourself a sweet cocktail using rum, you can ask Bing Chat, "What sweet cocktails can I make with rum?"

Bing Chat will then give you cocktail suggestions. If you like any, you can follow up on how to make it.

You can also use Bing Chat to generate unique drinks from fun prompts.

For example, at a Bing press event in NYC, the company had a smoothie bar that used Bing Chat to make smoothies according to your favorite TV character.

Also: This AI-powered app lets you chat with historical and fictional characters

The smoothie bar attendee typed a prompt into Bing Chat like, "Make a smoothie inspired by Dora," to which Bing outputted a detailed, fun recipe for incorporating traits of the character that the smoothie maker would follow.

You could do this with many other prompts, including seasons, feelings, songs, and more.

4. Date nights

If you have ever been in a relationship, especially a long-term one, you may have found yourself at a loss for new date ideas and feel like you've done everything.

AI chatbots can help you come up with fun, new, creative dates that you may have never thought of on your own.

Also: ChatGPT: Who's using the AI tool and why?

For instance, I asked Bing Chat for help coming up with the perfect, laid-back date night by asking, "Can you give me a date idea that is indoors and low budget?"

Bing Chat then gave me five cute date ideas that fit the criteria. Even if you don't decide to go with any of those ideas exactly, it could be a helpful way for you and your partner to get the creative juices flowing.

5. Finding what to watch on TV

Deciding to watch something on TV is only half of the battle. The trickiest part can oftentimes be deciding what to watch. Next time you are stuck, ask AI for help.

Also: The best Netflix alternatives

You can ask an AI chatbot to help you find a movie or show for you to watch in the specific genre you are looking for. It will make content suggestions that narrow down your choices and hopefully make your decision easier.

If you ask ChatGPT the question, it even includes a synopsis of each movie it is suggesting so that you can avoid watching the trailers. The only problem is that the suggestions would be limited to movies prior to 2021.

Also: The best video streaming services

The perk of using Bing Chat is that you can specify the streaming network you would like it to suggest movies from, too.

For example, I asked, "Can you help me pick a rom-com to watch on Netflix?", to which I was met with an entire list.

6. Play games

Although it may sound silly, playing basic games with AI is pretty fun. If you are ever sitting in a boring class or work meeting, feel free to ask your favorite AI chatbot to play a game with you.

AI chatbots are capable of playing all of the standard games you know and love, such as tic tac toe, hangman, trivia, 20 questions, and more.

Also: Can you distinguish people from AI bots? 'Human or not' online game reveals results

My personal favorite is tic-tac-toe, even though I found beating the AI a lot tricker than you would anticipate.

To get started all you have to do is say, "Can we play [insert game you want to play]" and the game will begin.

7. Plan a party

Planning a party involves so many moving pieces. Now you can use AI chatbots to help you with nearly every piece of the puzzle. To show you what I mean, I will start with an example.

Also: The best AI art generators

I asked Bing Chat, "Can you help me plan my dog a 5th birthday party?"

The initial response included venue, decorations, activities, and food ideas that I could use for my dog's birthday party. From there, I could move forward with any of those aspects and ask Bing Chat for more help.

If I was actually throwing this party, I could have followed up that question with questions regarding a more robust list of activities, treats, venues, food options, and more.

You could even use an AI image generator to help you create fliers, invites, and table placements.

Artificial Intelligence

YC-backed Dumme raises $3.4M for its AI video editor that turns long-form YouTube videos into Shorts

YC-backed Dumme raises $3.4M for its AI video editor that turns long-form YouTube videos into Shorts Sarah Perez @sarahintampa / 7 hours

Dumme, a startup putting AI to practical use in video editing, is already generating demand before opening up to the public. The Y Combinator-backed company has hundreds of video creators testing its product, which leverages AI to create short-form videos from YouTube content, and a waitlist of over 20,000 pre-launch, it says. Using a combination of both proprietary and existing AI models, Dumme’s promise is that it can not only save on editing time but also — and here’s its big claim — do a better job than the contracted (human) workforce who is often tasked with more menial video editing jobs, like cutting down long-form content for publication on short-form platforms like YouTube Shorts, TikTok or Instagram Reels.

Founded in January 2022 and a participant in startup accelerator Y Combinator’s Winter 2022 program, Dumme co-founder and CEO Merwane Drai said he was originally focused on building a search engine for video. But around six months ago, the team realized that a better product might be to repurpose the same AI models they were developing to edit video clips instead.

Joined by co-founders Will Dahlstrom (CPO) and Jordan Brannan (CTO), all with AI backgrounds, Drai realized Dumme may have landed on the right product-market fit after their app went viral, crashing their servers.

“We didn’t really expect that it would get a lot of traction or anything, so we just put something out there,” Drai explains. “Then what happened is that overnight, we woke up to overloaded servers — like, nothing actually worked. So we took everything down and actually put together some sort of waitlist,” he continues. “The next morning, we probably woke up to 5,000 people in there, which was interesting.”

The team later discovered that a TikTok creator had posted a short video about the product, which sent a flood of traffic to their site.

“It actually never calmed down from that,” Drai notes.

The product, pronounced “dummy,” appealed to creators because it aimed to simplify and speed up the work involved with video editing.

Image Credits: Dumme

Using Dumme is as simple as the name implies. To get started, the user pastes in a YouTube video’s link, then clicks on “generate” and the AI will output a number of short videos showcasing highlights from that ingested content. The company says it’s using YouTube as the source, instead of supporting raw video footage, in order to outsource content moderation — that is, if it’s allowed on YouTube, it’s allowed in Dumme.

The processing time and the number of resulting clips will depend on the length of the original video.

But as an example, an hour-long video podcast might take around 20 minutes to process and you’ll start receiving clips after about five minutes, Drai says. When complete, creators can download the video clips, which are under 60 seconds by default, and upload them to any platform that supports short-form content, like YouTube Shorts, but also other platforms, like Reels or TikTok.

Image Credits: Dumme

How this all works on the backend, of course, is much more complex. The company says that, initially, Dumme will learn as much as it can from the source video via the metadata. It then transcribes the video and tries to understand the semantics of what’s being said while also looking at the frames to try to decode the emotions of the person speaking. These findings are correlated and passed to a language model that tries to determine what parts of the video are worthy of clipping. That’s then handed off to another model that tracks active speakers and handles cropping.

Dumme says it’s working with existing AI models like GPT-4, a fine-tuned version of Whisper, and others it built in-house — like the model that tracks the active speakers in a video frame. One of its models is also trained on a bunch of YouTube Shorts to learn what makes for a good opening hook to pull viewers in. And, though not yet live, the team is also experimenting with an open source model, LaViLa from Facebook Research, to better understand the context of the video.

The AI work is being done on GPU Cloud provider CoreWeave, not AWS, as it’s more affordable, the company tells us.

Because Dumme relies on AI that processes spoken words, the tech is not appropriate for things like long gameplay videos or others where people aren’t talking. Drai says the startup is initially targeting YouTube creators, podcasters, and agencies — the latter, they believe, would be the best bet for monetizing the product.

Image Credits: Dumme

Agencies, explains Drai, today often outsource this type of work with hit-or-miss results.

“They just pay contractors in cheap jurisdictions to edit their own content. And the problem is that it’s still actually pretty expensive and it takes a lot of time — it takes weeks, not minutes,” he says.

Asked how he feels about creating a technology that would actually put people out of work, Drai was not worried.

“The way I think about it is that, eventually…I think this is like telling me that math teachers are going to [be put] out of work because there’s something called a calculator…,” he explains. “People are going to adapt. And then there’s going to be someone teaching you about the calculator, right? So I think it’s just a matter of adapting to this,” Drai says.

Currently, the pricing being considered involves tiers where a business would pay $0.40 per minute of video processed, while smaller creators may instead opt for a monthly subscription that’s capped at 10 hours of content per month. (These numbers may change.) During testing, the product has been free to use.

Early adopters have been using Dumme for a variety of edits, including generating clips from their video podcasts to publish to Shorts, as well as cutting down other new videos and going through their back catalogs.

The product appears to be competitive with other AI technologies on the market, including that from creator company Jellysmack, which has been leveraging AI to turn longer YouTube videos into shorter videos, by cutting them, resizing them, and optimizing them for specific platforms — a result of its 2021 Kamua acquisition. Other tools doing similar work include things like Opus Clip, Vidyo.ai, Detail, TubeBuddy, Wisecut, and others. To what extent Dumme succeeds or fails will be on outperforming competitors on the quality of work and cost — metrics that are yet to be determined.

But some investors are placing their bets on Dumme. Ahead of launch, the startup raised a $3.4 million seed round from Y Combinator, Caffeinated Capital, Max and Nellie Levchin (through SciFi VC), Suhail Doshi, Nico Chinot, Protocol Labs, Chris Puscasiu, and other angels.

Given the interest and the sizable waitlist, Dumme says it’s aiming to onboard around 500 people every week. TechCrunch readers can jump the line using the invite code TECHCRUNCH until the slots run out.