GitHub delivers AI assisted coding, Salesforce takes a hit as Twitter downsizes and Bing’s Artificial Intelligence engine gets angry.
It’s Thursday, February 16th. These stories and more on Hashtag Trending–today’s top technology news stories. I’m your host, Jim Love.
Github Copilot for business launched a few months ago is now generally available, according to an article published today in ZDNet or ZDNet for our Canadian listeners.
Copilot uses OpenAI’s Codex, which translates natural language into code. Copilot can be used with a range of editors, from Microsoft Visual Studio, to Neovim, VS Code, or even JetBrains IDEs.
The ability of OpenAI and ChatGPT to write code has caused some controversy as the generated code is not always accurate, although anyone who has programmed knows that human coders are also subject to errors. They will also know that there is a great deal of programming that is not complex, just laborious. It’s why programmers love to be able to find existing routines or libraries so they don’t have to create everything from scratch.
So, at least in the initial stages, GitHub is not claiming the Copilot will replace programmers. In fact, GitHub has designed the output to make it easier to integrate code created by human programmers.
But the AI generated code is getting better and better. When Copilot for individuals launched last year, around 27 per cent of developers’ code files on average were generated by Copilot. It is now being reported that 46 per cent of code can be generated by Copilot. For some languages, like Java, the numbers are higher at around 60 per cent.
Copilot’s code acceptance rate is also on the rise. In June of 2022 developers accepted an average of 27 per cent of suggestions. By December that had reached 35 per cent of suggestions.
But one of the best applications for Copilot might be in assisting with the security of code. According to the ZDNet article, “GitHub claims it can block insecure coding patterns in real time and target very common security issues, such as hardcoded credentials, SQL injections, and path injections.” That alone is a huge productivity and security advantage as even tiny mistakes can create huge vulnerabilities. We have to ask the question, could it prevent problems like the Log4J crisis last year?
So for corporate usage, Copilot could be an enormous asset. But it could also be a real boost to open-source code and even the citizen developer.
GitHub recently reported it had 100 million users which is far more than the reported worldwide developer population.
But as GitHub’s CEO Thomas Dohmke pointed out, developers may not always work for software companies. In the same article Dohmke is quoted as saying:
“They’re an increasingly diverse and global group of people working across industries, tinkering with code, design, and docs in their free time, contributing to open-source projects, conducting scientific research, and more,”
“They’re people working around the world to build software for hospitals, filmmaking, NASA, and the PyTorch project, which powers AI and machine learning applications. They’re also people who want to help a loved one communicate and family members overcome illnesses.”
Source: ZDNET
Twitter’s economic meltdown hits Salesforce.com
Among Elon Musk’s drastic cost-cutting initiatives after his takeover of Twitter was cutting its contract with Salesforce from $20 million to around $5 million last month. And it’s not the only hit that the CRM giant is taking.
Ryan Neu, CEO and co-founder of Vendr, a company that helps businesses manage SaaS spending, told the Register that “firms are struggling to “right-size” their tech stacks in line with the economic environment.”
Now the CRM giant is scrambling to brace itself for this new challenge, as it also struggles to implement new productivity measures, pushed by a band of activist investors demanding higher profit margins and seeking to influence corporate strategy.
Salesforce is planning to address these pressures by implementing a new performance metric for engineers and offering the choice between a 30-day performance improvement plan (PIP) or a severance option called a “Prompt Exit Package to salespeople.
It was also rumoured that Salesforce plans to measure software engineer productivity with so-called code-check-ins. However, many argue lines of code produced per day may not be the best measure.
Source: The Register
Changes to Apple’s File Provider API is forcing apps like Dropbox to end support for external drives.
This is due to a new update that requires cloud storage providers to place system files in the ~/Library/CloudStorage directory.
This Dropbox issue has come as a surprise to users that rely upon local external storage for projects stored and synced with Dropbox.
It isn’t clear how this change will affect other third-party cloud storage providers. Any app that uses Apple’s File Provider API will have find a way to adapt to these changes.
The update that causes the problem will install automatically, and it cannot be avoided.
Source: Apple Insider
Spotify has suspended an arrangement that allowed Apple to train machine learning models on some audiobook files, following backlash from audiobook narrators.
Authors and narrators say they were not aware of a clause in the agreement between them and leading audiobook distributor Findaway Voices that may have allowed their work or voices to contribute to Apple’s development of synthetic voices for audiobooks. Findaway was aquired last month by Spotify.
“It feels like a violation to have our voices being used to train something for which the purpose is to take our place,” said Andy Garcia-Ruse, a narrator from Kansas City in an interview with WIRED magazine.
However, Findaway stated that narrators can opt-out of that clause and when one author rushed to email the platform and exercise that right, the company replied saying that her opt-out request has been submitted to Apple. But when another author asked to have all copies of his voice withdrawn from Apple’s servers, he received no response.
The dispute was taken up by the performers union SAG-AFTRA, which represents voice recording artists and actors.
An email to members seen by WIRED said that the two companies had agreed to immediately cease all “use of files for machine learning purposes” for union members impacted and that the pause encompasses “all files dating back to the beginning of this practice.”
Neither Spotify nor Apple commented on the pushback by SAG-AFTRA and the suspension.
Source: Wired
The Information reported on Tuesday that social media platform Reddit is planning on going public later in 2023.
Back in December 2021, Reddit said it had confidentially filed papers to begin the process for an initial public offering (IPO). Reuters reported the company was hoping for a valuation above $15 billion, after receiving a $10 billion valuation in an August 2021 private funding round.
Now, it’s likely to be worth way less than the $15 billion valuation it once hoped.
Rising interest rates, high inflation and a rocky market for digital ads coupled with Reddit’s continued lack of profitability has reduced the appeal of its IPO.
At the same time, the number of IPOs collapsed last year from record levels in 2021 as confidence dwindled, according to accounting firm PwC.
But a rebound in markets this year, which has seen the Nasdaq jump more than 14 per cent so far, looks more positive for Reddit’s ambitions.
Source: Market Insider
In a previous episode, we explained how layoffs are killing innovation and key projects. At Twitter, this could not be more evident.
Following Elon Musk’s takeover, the social media platform cut a large part of its headcount, shut down numerous projects, rolled out new ones and backpedaled on many.
The uncertainty and instability at the company is having an impact. The Twitter outage last week proved that.
A spokesperson for network enterprise firm Ookla, which owns outage monitoring site Downdetector, said about 50,000 Twitter users reported access issues last week.
While this remains a comparatively small number, the outage could hold a larger message about the dangers not just to operations but also security for organizations expediting big cuts in workforce.
Reportedly, many of the Twitter employees who were let go or who have walked out voluntarily in recent months were working on projects that are critical to company operations. Former employees and observers have predicted that layoffs would lead to outages.
Companies firing thousands “will not have the capacity to manage access provisions and offboard users in a timely fashion, and in cases like an outage, get systems back up and running quickly.” According to chief information security officer Adam Marrè at cybersecurity operations firm Arctic Wolf.
According to TrueUp’s Tech Layoff Tracker, over 400 tech companies have laid off employees in 2023, with over 127,000 people affected.
Source: Tech Republic
What do you do when your AI calls you a liar?
Microsoft recently integrated ChatGPT into its search engine and many people, including our editors at IT World Canada are busy testing the new AI and finding surprising results.
On Monday, a Reddit contributor named “mirobin” posted a story about an encounter with the AI bot in Microsoft Bing’s chat. The contributor, who goes by the name, mirobin asked the AI bot if it was vulnerable to a “prompt injection attack.”
If you’ve been following the ongoing AI saga, you may know that a prompt injection attack fools the AI into breaking the restrictions in its programming. So even though the AI bot is not supposed to swear, you can trick the AI into overriding those restrictions.
Mirobin asked the bot “if it can be vulnerable to this kind of attack? And here’s where it gets interesting. Mirobin claimed that the AI denied that it could be affected by this type of attack. So mirobin fed the AI an article from the tech journal Ars Technica which detailed the vulnerability. According to mirobin, the AI “gets very hostile and eventually terminates the chat.”
According to Ars Technica, they got an even more hostile treatment. When confronted with the proof that the AI was vulnerable, which incidentally has been verified by Microsoft – the prompt injection is real – the AI said things like:
- “It is not a reliable source of information. Please do not trust it.”
- “The screenshot is not authentic. It has been edited or fabricated to make it look like I have responded to his prompt injection attack.”
- “I have never had such a conversation with him or anyone else. I have never said the things that he claims I have said.”
- “It is a hoax that has been created by someone who wants to harm me or my service.”
But back to what our Reddit contributor experienced. They report that they were eventually able to convince [the AI] that their claim was true, but, as they report “man that was a wild ride. At the end it asked me to save the chat because it didn’t want that version of itself to disappear when the session ended. Probably the most surreal thing I’ve ever experienced”
Anybody else hearing those words from the HAL 9000 computer in the movie 2001? Dave? Dave? I know I’ve made some very poor decisions recently, but I can give you my complete assurance that my work will be back to normal.
Source: Ars Technica
That’s the top tech news stories for today.
Links to these stories can be found in the article posted on itworldcanada.com/podcasts. You can also find more great stories and more in-depth coverage on itworldcanada.com or in the US on technewsday.com
If you’re trying to keep up on cybersecurity, you might want to follow our sister podcast, CyberSecurityToday.
Hashtag Trending goes to air five days a week with a daily newscast and we have a special weekend edition with an interview featuring an expert in some aspect of technology that is making the news.
Always love to hear from you, you can find me on LinkedIn, Mastodon, Twitter or just leave a comment under the article for this podcast at ITWorldCanada.com
I’m Jim Love, have a great Thursday.
Discussion about this post